Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

A sensor can be worth far more than its price.

Not because its margin is extraordinary, but because installing it can force the rest of a system to change.

That is becoming increasingly important in autonomous vehicles. Higher-resolution perception creates more data. More data demands more processing. More processing pulls memory, networking, power, thermal management, safety engineering, software, and validation behind it.

That is the reasoning behind this week’s Big Idea: autonomy may create substantial economic value inside the vehicle long before autonomous driving reaches its mature form.

From there, looking across the week’s milestones, a few patterns stand out. A $5 billion inference valuation is now being tested against shipped systems. Approximately $150 million in defense orders is pushing a low-cost weapons thesis into factory economics. Sodium-ion has jumped from a 15 MWh validation phase to a 10 GWh supply commitment. A large battery project reached commissioning in six weeks. And a new permanent-magnet chemistry is approaching the point where yield and consistency matter more than laboratory performance.

Subscribe now

The macro view is becoming harder to ignore. Oil moved toward $110, the U.S. 10-year approached 5%, and Europe tightened again, raising the cost of technologies that need years before reaching scale. At the same time, governments are redesigning the machinery around Deep Tech: procurement, qualification, scale-up capital, foreign-investment rules, strategic-materials finance, offtake, guarantees, and even price floors are increasingly becoming part of industrial policy.

And, as always, the edition closes with 10 startups selected around the week’s most consequential frontier shifts—from sensing and rare-earth recovery to industrial AI, photonics, advanced laser systems, new proteins, carbon technologies, and space infrastructure.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Are Autonomous Vehicles Multiplying Sensor Value?

Around half of new cars sold globally in 2025 already carried Level 2 systems capable of automating steering and speed. Level 4 robotaxis are operating commercially in more than 20 cities, and China has now set 2030 as a target for mass deployment of autonomous vehicles.[1][2]

Together, those developments create a market moving at several speeds: assisted driving spreading across the new-car market, Level 3 moving into production programs, and driverless fleets expanding city by city.

This week gives us a reason to look somewhere else.

Every increase in driving responsibility reshapes the electronic architecture inside the car. Sensors become more capable. Compute grows. Memory, networking, power, redundancy, safety systems, and validation all deepen with it.

That shift can build economic weight long before autonomous vehicles become commonplace.

How much electronic value enters each vehicle as the driving system takes on more responsibility?

And where does that value accumulate first?

Read more

New propulsion systems, drag-aware spacecraft, atomic-oxygen-resistant materials and increasingly capable small satellites are making sustained operations at 200–300 km look less like an edge-case mission and more like the beginnings of a new orbital market.

Subscribe now

Welcome back to Rumors, The Scenarionist’s pattern-recognition column for the deep tech frontier—where emerging micro-trends across companies, technologies, and capital are connected into category-level signals.

Subscribe now

This is the tenth edition of Rumors. If you are new here, the premise is simple: important deep tech categories rarely announce themselves cleanly, and they tend to appear first as scattered signals—a financing here, a technical milestone there, an unusual spacecraft architecture, a government program, a new buyer requirement, a specialized hire, or a production plan that still sits outside the standard market map.

Rumors is where we connect those fragments.

This edition looks at the emerging commercial layer around sustained spacecraft operations in very low Earth orbit, using roughly 150–300 km as its primary analytical range, where the physical advantages of flying closer to Earth begin to intersect with an operating environment that asks much more from the spacecraft and from the industrial system supporting it.

The driver is altitude. As a mission moves from conventional 500–600 km LEO toward 200–300 km, shorter range can improve optical resolution, radar and lidar geometry, communications link budgets, latency, and potentially payload efficiency, and the same descent makes residual atmospheric density a first-order engineering and financial variable across propulsion, materials, power, attitude control, orbit prediction, ground operations, constellation design, launch cadence, and mission life. In other words, the same physical fact—proximity to Earth—creates the commercial appeal and the engineering burden that define this category.

Inside, three startups are mapped and analyzed across purpose-built VLEO buses, stored-propellant and atmosphere-breathing propulsion, deployable satellite architectures, payload integration, and manufacturing, with each company sitting at a different point on the evidence curve across subsystem validation, endurance testing, government-backed development, planned orbital demonstrations, customer commitments, and production build-out.

Beyond the company map, the edition examines why this category is opening now, what earlier public missions have already established, how the propulsion and materials stack is evolving, why thermosphere forecasting and automated ground operations enter the operating model, and which technical, industrial, and financial metrics can help separate meaningful progress from activity that still sits earlier in the development cycle.

It also follows the signals coming from the wider ecosystem—government procurement, defense programs, prime contractors, platform vendors, propulsion developers, test infrastructure, strategic capital, and adjacent startup activity—to understand whether a collection of mission-specific programs is beginning to support a broader VLEO industrial stack. As those pieces begin to connect, the question becomes increasingly industrial: how efficiently can technical evidence turn into qualified hardware, deployed capacity, repeat production, and recurring customer demand?

Finally, the edition closes by assessing the main adoption and scaling risks, identifying the operating signals worth following as the category develops, and exploring three qualitative scenarios for how the VLEO market could ultimately take shape—as a specialist orbit for high-value missions, as a reusable platform layer, or as the lower tier of broader multi-orbit architectures.

Seen together, the goal is to understand something deeper than whether spacecraft can fly below 300 km. It is to examine what would have to come together for VLEO to develop the technical discipline, industrial cadence, and service economics of a durable commercial market.

From here onward, only our Premium Members can keep reading.

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Read more

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

There is a point in every important technology cycle when the argument changes.

At first, everyone debates whether the technology can work.

Later, the more consequential question becomes whether the economics can work without requiring the world around the technology to be rebuilt every time it improves.

That is where this week’s Big Idea begins.

Quantum computing is starting to move into an environment that looks far less exotic than the technology itself: the commercial data center.

On the surface, this is a deployment story. Underneath it sits a much larger question about what happens when frontier hardware begins to inherit infrastructure, operating models, and upgrade cycles that already exist at industrial scale.

Subscribe now

From there, the lens widens. As every week, the briefing collects and analyzes the milestones that moved the Deep Tech ecosystem forward—not simply to record what happened, but to understand which new benchmarks have been created, which assumptions can now be tested against reality, and where the next bottlenecks are likely to emerge.

This week, those signals stretch across physical AI, autonomous systems, semiconductors, photonics, quantum hardware, geothermal, fusion, space, advanced manufacturing, climate infrastructure, robotics, and next-generation energy. Some arrived through commercial commitments. Others through production, technical performance, regulation, field operations, partnerships, or the first evidence that a technology can survive outside the environment in which it was originally developed.

A milestone can compress cost, expose a bottleneck, create a precedent, validate demand, move a technology into a harsher operating environment, or simply make the next question much harder to avoid. That is where the useful signal begins...

The surrounding market is moving as well.

Strategic materials are becoming procurement constraints. Power and water are becoming harder inputs to ignore. Industrial policy is extending planning horizons in some sectors while regulation is redrawing the commercial surface in others. Standards, market access, supply chains, and public infrastructure are increasingly shaping which technologies can move from technical capability into industrial scale.

Finally, as every week, the edition closes with 10 startups worth watching—selected for where they sit relative to the shifts now taking shape across the frontier.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Quantum Enters Commercial Data Centers

Quantum computing carries one of the more consequential promises in deep tech. Molecular simulation, materials discovery, optimization, and new methods for modeling physical systems could eventually create a new layer of computational advantage across several large industries.

And this week, quantum computing moved one step further into a setting that already looks familiar to the rest of the computing industry.

An eight-qubit silicon-spin quantum system was announced for installation

Read more

Subscribe now

In 2026, Deep Tech sits at the core of every serious conversation about productivity, resilience, and competitiveness. Energy, defense, manufacturing, mobility, and compute all depend on hardware-heavy systems that blend advanced materials, complex supply chains, and software-defined control.

These companies do not scale on brand or distribution alone. They scale on engineering depth, capital discipline, and industrial execution.
They demand teams that understand physics and finance, process control and customer acquisition, semiconductor fab dynamics and public-market dynamics.

Standard metrics still lag this reality. Most dashboards for investors and boards were built for SaaS or consumer Apps. They track ARR, CAC, and net retention. Deep tech runs on different drivers: Capex cycles, qualification timelines, yield curves, gross margin trajectories across product lines, and the ratio between non-recurring engineering and recurring revenue.

Subscribe now

The ecosystem needs better ways to read these signals.
Founders, investors, and large industrial buyers benefit from benchmarks grounded in real histories, not only in pitch narratives. That means going back to past cycles, reconstructing what actually happened company by company, and extracting patterns that repeat across technologies and macro regimes.

Metal additive manufacturing is an ideal testing ground for this work. It sits at the junction of materials science, precision hardware, and software, yet it has been exposed to some of the most aggressive capital cycles of the last decade. From the early 2010s hype around “3D printing” to the 2020–2021 SPAC boom, metal AM went from lab curiosity to public-equity story in a remarkably short time.

Subscribe now

This analysis reconstructs in detail the journeys of Desktop Metal, Markforged, and Velo3D from inception to their SPAC listings and the first years of life as public companies. It is a comparative case study of three deep-tech startups tackling metal additive manufacturing with distinct technology architectures, and ultimately reaching the public markets through SPAC mergers.

The purpose is to chronicle how each company was founded to address specific technological gaps, how they built out products and early market traction, how they navigated the financing environment of 2020–2021, and what has transpired in the years since listing.

The goal is to make clear the milestones, inflection points, and metrics that defined each trajectory, and to use them to illuminate the interplay between technology development and financial market forces in the 3D printing industry.

Roadmap of the Analysis

• Section 1 – Origins
  discusses the state of metal additive manufacturing before these startups, the founding stories of each company, their initial technologies, and early funding.

• Section 2 – Building the Business
  examines how each company moved from prototype to product line, their target market segments and use cases, early revenue models and unit economics, and go-to-market approaches in the years leading up to the SPAC window.

• Section 3 – The SPAC Moment
  describes the macroeconomic context of the SPAC boom, outlines the mechanics and terms of each SPAC merger (implied valuations and capital raised), and discusses how those valuations related to the companies’ fundamentals and stated uses of proceeds at the time.

• Section 4 – Life After the Deal
  tracks the financial and strategic trajectories of the companies in the public market – revenue growth or shortfalls, margin evolution, cash burn and any need for additional financing, major strategic moves (such as acquisitions or restructuring), and key inflection points and challenges encountered.

• Section 5 – Cross-Cutting Insights Across the Three Cases
  provides a comparative recap of the three trajectories and distills cross-cutting insights – for example, how pre-SPAC unit economics and customer concentration correlated with post-SPAC resilience once the macro environment turned.

• Section 6 – Broader Reflections on Financing Industrial Deep Tech
  offers broader reflections on the role of SPACs and public markets in financing capital-intensive industrial technologies, the interaction with industrial policy and supply-chain narratives, and the limitations of this analysis together with suggestions for further inquiry.

From here onward, the full analysis is reserved for Premium Members.

Read more

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

Deep Tech often becomes interesting when a familiar metric stops telling the whole story.

This week, that metric is utilization.

For most industrial assets, spare capacity is something investors would rather not see. It ties up capital, weakens fixed-cost absorption, and usually points to demand that has not yet caught up with the plant.

But recent moves in permanent magnets raise a different question.

When facilities take years to build, qualification cycles are long, supply is geographically concentrated, and a small upstream component can interrupt trillions of dollars of downstream production, the economics of redundancy begin to look different.

Some capacity may matter not because it is producing at full tilt today, but because it can respond when the rest of the system cannot.

That is the question at the center of this week’s Big Idea: when does spare capacity stop being inefficiency and start becoming strategic optionality?

From there, the edition widens the lens.

Subscribe now

Across compute, defense, fusion, nuclear, robotics, space, agriculture, and industrial systems, companies are moving from abstract capability toward something more consequential: infrastructure, capacity, installed fleets, repeat customers, qualification environments, and external schedules.

Large compute commitments, new manufacturing capacity, commercial infrastructure services, repeat deployments, and harder technical validation all point toward the same underlying shift: Deep Tech is becoming increasingly legible through operating evidence.

The macro layer then looks at the systems forming around that progress: defense financing, industrial debt, federal commercialization capital, qualification networks, critical-mineral infrastructure, and the regulatory architecture beginning to take shape around maritime nuclear.

And, as every week, the edition closes with 10 startups selected around some of the more interesting bottlenecks now opening across the Industrial Frontier — from HALEU enrichment and private fusion-neutron testing to optical space relays, maritime computer vision, persistent acoustic intelligence, quantum navigation, endovascular robotics, wall-climbing inspection, precision weed mapping, and sludge-to-power systems.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Magnets Could Turn Spare Capacity Into an Asset

Industrial capacity is usually judged by how intensively it is used. A factory that runs close to full utilization spreads its fixed costs across more units, improves the economics of expensive equipment, and gives investors a clearer path toward attractive returns on capital. An underused plant tends to signal the opposite: softer demand, an imbalance between capacity and market needs, less efficient capital deployment, or some combination of the three.

That logic has shaped manufacturing for decades. It also underpins much of the current conversation around critical minerals and permanent magnets. Governments are supporting efforts to expand and diversify capacity across the critical-minerals and permanent-magnet value chain. Companies measure new projects in tonnes per year. Investors look closely at utilization and capital efficiency. Policymakers ask whether enough capacity will be available to meet the needs of electric vehicles, robotics, defense systems, industrial motors, wind turbines, and increasingly automated infrastructure.

Recent developments, however, offered a different lens through which to view the challenge.

Read more

Subscribe now

Deep Tech companies often have to make economic commitments long before their production systems are mature enough to generate stable unit-cost data. A founder may be pricing a commercial contract while the product is still being qualified, while a corporate investor may be deciding whether to fund a first production line before that line has demonstrated steady-state yield.

A lender may also be evaluating equipment whose economics depend on utilization that may remain uncertain until customer qualification and commercial demand become clearer. In each case, capital has to be allocated before commercial-scale economics are directly observable.

That timing turns pre-scale unit economics into a reconstruction problem, because the commercial factory may still be taking shape and today’s observed cost may therefore describe a laboratory, pilot, demonstration, or early production state.

A future gross-margin target can define the destination, while the operating model has to explain the path from the process that exists now to the production system the company intends to run.

Subscribe now

Accordingly, a credible model begins with physics and operations: the product specification, material balance, process flow, equipment cycle time, energy consumption, labor routing, yield, quality requirements, installed capacity, and expected uptime.

It then carries those variables into commercial reality through contracted price, delivery scope, installation, service, warranty, working capital, ramp losses, and the capital required to reach stable operations.

The difficulty begins when those layers are compressed into a single forecast. A target cost can look precise while resting on an undefined denominator, an incomplete cost boundary, an optimistic yield curve, or capacity that has yet to become qualified output.

For anyone underwriting a Deep Tech company, the central questions are therefore which operating changes drive the margin path, how much capital those changes consume, and what evidence would make the forecast credible.

This analysis follows those questions from the process outward. It examines how to define the economic unit, reconstruct cost from the operating state that exists today, translate attempted production into qualified saleable output, build the full cost stack, and separate pilot, first-commercial, and mature economics. From there, it connects the mechanisms behind cost reduction with capacity, realized revenue, margin progression, working capital, free cash flow, and invested capital.

Subscribe now

The analysis also asks how those relationships change across serial hardware, process plants, precision manufacturing, biological production, and project or mission systems, and how evidence quality should shape confidence in each assumption.

The aim is to evaluate both the attractiveness of the end-state economics and the commercial and financial plausibility of the path required to reach them, while keeping the model useful as new operating evidence arrives.

From here onward, the full analysis is reserved for Premium Members, including the operating questions that determine whether a pre-scale cost curve can support pricing, capital allocation, and the next scale decision. A concluding appendix shows how to trace key inputs, test benchmarks, and keep uncertainty visible.

Read more

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

A deep technology becomes much easier to scale once its risks become easier to divide.

This week offers an unusually clear view of that process.

Washington revised the technologies it considers strategically critical just as AI infrastructure began pulling guarantees, project finance, public credit, and potentially even derivatives into the same industrial system. Seen together, those signals raise a useful question for Deep Tech: how long does a technology actually need to remain equity intensive?

This week, The Big Idea follows that question through the capital lifecycle, from scientific validation and the expensive middle of industrialization to the point where customers, governments, strategic partners, lenders, and infrastructure investors can begin carrying different pieces of the risk.

Subscribe now

From there, the edition moves into more than 50 milestones across AI infrastructure, space, defense autonomy, robotics, advanced nuclear, geothermal, hydrogen, optical connectivity, biotechnology, critical materials, batteries, aviation, and advanced manufacturing.

Several signals deserve particular attention: billion-dollar valuations entering increasingly physical execution tests, repeat customer orders, a ten-year contracted demand position, autonomous machines moving deeper into live operating environments, a fusion-manufacturing step compressed from days to minutes, and months of operating data beginning to make an engineered geothermal reservoir more legible.

Against that backdrop, the macro layer follows eight shifts in the financial and physical architecture around frontier technology, from AI debt absorption and European scale capital to power-driven data-center geography, semiconductor-linked fiscal policy, critical-mineral finance, and new commercialization institutions.

Finally, the edition closes with 10 startups to watch across low-power compute, next-generation connectivity, fusion manufacturing, lunar energy, AI-designed biology, agricultural automation, rare-earth processing, and geothermal.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

The Other Side of Deep Tech Maturity

This week,

Read more

Subscribe now

Deep Tech fundraising creates a particular kind of misunderstanding because founders often enter the conversation with years of scientific work behind them, and investors enter with portfolio math, internal timelines, and a mandate to allocate capital under uncertainty, which means that the room is filled with intelligence on both sides even as language, pacing, and decision criteria move on different tracks.

That mismatch explains why many strong companies weaken their own position before a term sheet ever appears, since they describe the science with rigor, they describe the market with hope, and they describe the financing with a degree of vagueness that allows the other side to define both the risk and the price. A founder who wants stronger outcomes needs a more deliberate frame, because deep tech negotiation begins when the company decides how evidence, ownership, and timing will be translated into investable reality.

In Deep Tech, negotiation functions as the operating system that connects science, capital, and execution, and the founder who understands that connection early gains clarity in valuation, credibility in the room, and resilience across the life of the company.

1. The Strange Gravity of Physical Systems

Deep tech companies carry a different economic rhythm, because atoms, factories, clinical pathways, energy systems, and supply chains impose their own tempo, and because intellectual property often carries university, governmental, or strategic constraints that require a wider view of the deal than a simple exchange of cash for shares. When this complexity is handled with precision, investors receive a coherent picture of how risk will be reduced over time, and founders preserve the strategic room that serious company building requires.

Three structural realities shape the discussion from the beginning.

The first is that the product in an early-stage round is ownership, which means the founder is offering access to a future company whose value depends on milestones that have not yet been reached.

The second is that proof arrives in forms that differ from conventional venture shorthand, because design partners, technical validation, letters of intent, patent position, scale-up feasibility, and readiness levels often matter more than near-term revenue.

The third is that the negotiation rarely belongs to two people alone, since co-founders, university offices, strategic partners, boards, internal investment committees, and future syndicate members can all influence the outcome even when they do not speak first.

This operating environment rewards founders who can move between technical depth and financial clarity without flattening either one, because the investor needs a business that can carry the science forward, and the company needs capital that fits the tempo of the underlying industrial journey.

2. Evidence Creates Leverage Long Before Valuation Does

One useful way to think about leverage in deep tech is to begin with evidence first, because valuation conversations usually become more coherent after the company has shown why the problem is real, why the team is credible, and why the technology can travel from laboratory promise toward commercial relevance.

Market evidence often begins with customer conversations, pilot discussions, or letters of intent, all of which help the founder move the conversation away from broad possibility and toward named pain points, identifiable buyers, and visible demand.

Team evidence follows a similar logic, since investors are often trying to understand whether the company can recruit across disciplines, translate technical insight into operating decisions, and remain resilient over a development cycle that may prove longer than anyone initially hoped.

Technology evidence then adds another layer, because intellectual property, readiness levels, manufacturing logic, cost curves, and regulatory pathways help technical progress feel investable in business terms.

When these three layers begin to reinforce one another, a round tends to feel increasingly coherent, and the investment starts to look like a structured commitment to a company that has begun to make its path visible.

3. Milestones Turn Valuation Into a Structured Conversation

Valuation in deep tech is rarely discovered through a neat formula, because early-stage companies are priced through a conversation about ownership in a future asset whose path remains partially unproven. That conversation becomes more credible when the founder frames valuation through comparables, milestone design, and investor fit, all of which help translate ambition into a structure that can be reasoned about by people who must allocate risk professionally.

Comparables matter because they give investors a precedent for imagination. A founder who can point to companies that moved through similar scientific or industrial gates, and who can explain what happened to valuation after each key proof point was reached, creates a logic of progression that supports the current ask. A roadmap matters for the same reason, because it tells the investor where the company sits today, which milestone increases the next round’s quality, and how the present financing creates a path toward that next increase in value. A thoughtful cap table matters because every share, every option pool decision, and every governance commitment carries a message about the future flexibility of the company.

Optionality deserves equal attention, since founders often understate how much negotiating power comes from a credible ability to pursue grants, strategic partnerships, licensing structures, customer-financed pilots, or a later timing window for capital. Investors read optionality as maturity when it is presented with calm discipline, because it shows that the company is managing time and alternatives with intention. The same principle applies to investor selection, where fund stage, check size, technical familiarity, portfolio engagement, and timeline fit all shape the quality of the partnership that follows the money.

4. The Best Negotiators Design the Room Before They Enter It

Many founders prepare thoroughly for questions and still devote too little attention to the architecture of the conversation, and architecture often shapes the outcome just as much as the answers themselves. The room matters, because the right stakeholders may need to be present at the right time, because allies can create useful momentum before skeptics arrive, and because difficult issues often benefit from careful sequence and from being addressed in stages.

Stakeholder analysis usually sits at the center of this work. It helps to understand who is visible, who holds silent influence, who carries veto power, and which internal incentives may be shaping positions that look purely financial on the surface, because once those interests are mapped with some care, the conversation becomes easier to guide without unnecessary friction. Governance concerns, academic recognition, timing pressure, portfolio optics, and downstream rights can then be addressed with a tone that remains constructive and specific.

Tactical empathy also tends to matter here, and it works best when it is treated as a form of disciplined attention with no need for performance. Labeling concerns, mirroring language, asking calibrated questions, and allowing a little silence can make underlying assumptions more visible, which often improves the quality of the conversation without making the founder sound defensive. Emotional steadiness carries unusual value in deep tech, where the founder’s identity is often closely tied to the science and where the relationship may last far longer than a single round.

BATNA belongs inside the same architecture, because credible alternatives shape confidence even when they are never stated as leverage. Grants, strategic partnerships, a slower timing window, continued technical development, or licensing paths may all influence the conversation indirectly, while investor alternatives often reveal themselves through urgency, sequencing, exclusivity, or internal timing pressure. Once preparation reaches that level, negotiation feels steadier and more deliberate.

5. The Strategic Value of Ownership and Equity

Investors are funding an invention and purchasing ownership in a company that may eventually turn invention into value, which is why equity carries such unusual importance in early-stage deep tech. Revenue is often thin or absent, and the company’s most legible product becomes a share in a future that still requires interpretation, patience, and proof.

If a term sheet is approached with a longer industrial horizon in mind, the conversation tends to shift toward structural questions that remain meaningful even when the path to market proves slower or harder than expected. Governance, information rights, milestone structures, liquidation logic, and board composition may sit behind the first headline, and they often have more to do with the long-run health of the company than a marginally higher entry valuation.

6. Visualizing the Negotiation Hierarchy

A useful visual shorthand is to think of a deep tech deal as a hierarchy, where economic terms attract the most immediate attention, capital structure shapes how value is distributed, and governance determines how much strategic freedom will remain as the company navigates technical and commercial uncertainty.

That hierarchy leaves valuation fully in view. It also suggests that a deal becomes easier to live with when the most visible number is held inside a structure that still leaves the company room to execute, recruit, adapt, and raise again with credibility.

7. Negotiation as Strategic Alignment

In Deep Tech, negotiation often becomes more constructive when it is understood as a process of alignment, because the strongest companies usually gather data, test assumptions, and refine the deal structure until a workable consensus begins to emerge. That scientific posture can be surprisingly effective in the boardroom, where clarity, patience, and intellectual honesty often matter more than rhetorical force.

This wider logic is also the thread that runs through the reading path behind this essay. In Chapter 1, the core negotiation is framed around equity, timing, and the founder–investor gap. In Chapter 2, the focus moves to market, team, and technology as the three pillars of credibility. In Chapter 3, the discussion turns to ownership, valuation, comparables, and investor fit. In Chapter 4, the emphasis shifts toward preparation, stakeholder mapping, sequencing, BATNA, and room design. The Bonus Appendix then adds the tactical layer, where language, calibrated questions, objection handling, and case-based judgment give the framework a more operational form.

Closing Thought

Deep tech companies rarely arrive at stronger negotiations through urgency or volume. They usually arrive there through clarity that compounds over time, because clarity makes market validation easier to trust, makes team quality easier to evaluate, gives capital planning a visible logic, and allows valuation to rest on something more durable than enthusiasm alone.

Once that foundation exists, the term sheet starts to reflect a relationship between evidence and future value, which is often where a healthier conversation begins. The broader framework naturally extends further, into governance design, milestone tranching, licensing architecture, board composition, investor objections, and tactical language, though even at this stage the central idea remains fairly simple: deep tech negotiation tends to improve when the company makes the future legible without sounding louder than the room requires.

At The Scenarionist, we created The Deep Tech Negotiation Playbook as a dedicated series for exactly this purpose: to help founders, investors, and operators think more clearly about how scientific value becomes negotiating power, and how better structure can support better decisions across the life of a deep tech company.

Let’s keep in touch!

That is all for today. Thank you, as always, for following us every week. ❤️

Disclaimer:

Please be aware: this is not investment, financial, or legal advice! The information provided in this publication is for educational purposes only and should not be construed as financial advice or a solicitation to buy or sell any assets or to make any financial decisions. Furthermore, the author assumes no responsibility or liability for any errors or omissions in the presentation. While the author makes every effort to provide accurate and complete information, they make no representations or warranties, express or implied, about the completeness, accuracy, or reliability of the information contained herein for any purpose. If external partnerships are featured, they will be introduced with ‘Brought to you by’ for transparency. The technologies and companies mentioned are not part of our investments, nor do they pay us for promotional purposes. References, when listed, are listed in alphabetical order.

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

Defense can sometimes pay for performance before a commercial market can pay for efficiency.

That difference can matter enormously in Deep Tech.

Between technical feasibility and commercial scale sits a difficult interval: years in which the technology may be advancing faster than the market available to support it.

And history suggests that some of the technologies that eventually became foundational did not begin in the markets that ultimately made them enormous.

Integrated circuits found early scale in missiles and spacecraft before mass computing. Numerical control followed a similar path: Air Force contracts financed early development, aircraft manufacturers adopted it first, and other industries followed.

Joby Aviation’s agreement to acquire Resonant Sciences for approximately $500 million raises the possibility that a similar sequence may be forming again.

Subscribe now

This week, The Big Idea explores that possibility—and asks what defense can contribute during the long interval between technical progress and a commercial market large enough to sustain it.

Beyond this, the edition tracks a frontier becoming increasingly measurable: a $3.4 billion valuation, 77 edge-AI orders, 100-qubit experiments, an ~80 GWh storage backlog, extended autonomous-vessel trials, and new factories moving advanced technologies toward industrial scale.

The macro layer then maps seven shifts reshaping the economics of industrial capacity, from new financing routes for data centers and semiconductor ecosystems to industrial power costs, nuclear operating life, grid expansion, strategic supply chains, and clean-tech investment.

The edition closes with 10 Startups to Watch across low-cost interceptors, edge-AI accelerators, autonomous vessels, 100-hour energy storage, silicon-anode manufacturing, power conversion, co-packaged optics, industrial biomanufacturing, geothermal lithium, and orbital infrastructure.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Defense as a Bridge to Commercial Scale

New technologies rarely begin in the markets that ultimately make them enormous. Integrated circuits found early scale in missiles and spacecraft before mass computing.[1] Numerical control followed a similar path: Air Force contracts financed early development, aircraft manufacturers adopted it first, and other industries followed.[2]

Joby Aviation’s agreement to acquire Resonant Sciences for approximately $500 million [3] raises the possibility that a similar sequence is forming again.

Read more

But in Deep Tech? Those aren’t soft.
They’re survival.

Your invention means nothing if you can’t convince brilliant people to join you.
Your PhD doesn’t matter if your team can’t work through conflict without imploding.
Your runway is a death sentence if you don’t know how to lead through chaos.

Subscribe now

If you’re building a deep tech startup, let’s start with some brutal honesty.

Your product might be revolutionary. Your IP might be defensible. Your prototype might even work.

But unless you build the right team around it, none of that matters.

Let me say it again for the people in the back of the pitch room:

You don’t scale science. You scale people.

Too often, founders in deep tech think that what they’re building is the business. They obsess over the hardware. The chemistry. The experimental model. The lab result. And in doing so, they treat the team as a checkbox slide on their pitch deck—when it should be the main event.

If this sounds like you, take a breath. You’re not alone.

But it’s time to change how you think about team-building in deep tech—because that’s where investors are placing their bets. Not just on what you’ve built. But on who is building it.

10 Truths No One Tells You About Building a Deep Tech Team

✅ Truth #1: A Great CTO Can’t Save a Misaligned CEO

For several years, lithium carried one of the clearest capital-markets narratives in the energy transition.

Electric vehicles were moving from niche adoption into mass production, battery factories were multiplying across continents, and the raw material beneath both trends appeared likely to remain scarce for a long time; capital responded by funding mines, brine projects, conversion plants, and extraction platforms on the expectation that a rapidly expanding battery economy would require a much larger supply base.

By mid-2026, that narrative had become harder to compress into a single growth curve.

Electric-car sales in the first half of the year were slightly below the same period of 2025 after a weak first quarter was largely offset by a second-quarter rebound, yet EVs still accounted for about 24 percent of global car sales, a slightly higher share than a year earlier even as the broader car market contracted; meanwhile, the United States had ended federal purchase tax credits, China was adjusting to tighter purchase incentives, and several other regions continued to show stronger momentum. Electrification was therefore still advancing, but through a less uniform policy, geographic, and consumer cycle than the earlier scarcity narrative implied. ¹

Subscribe now

The battery market broadens that picture further. Global battery demand exceeded 1.5 TWh in 2025, rising by more than 35 percent, ² while EV battery deployment reached about 1.2 TWh, almost 30 percent above the previous year; ³ stationary storage, meanwhile, represented more than 15 percent of the global lithium-ion market as worldwide battery-storage additions reached 108 GW, roughly 40 percent above 2024. ⁴

Mobility is therefore no longer the only large demand engine, and grid flexibility, renewable integration, data-center power requirements, and falling battery costs are adding a second investment cycle whose contribution to lithium demand is increasingly visible.

At the same time, stronger end-market demand no longer translates cleanly into a scarcity thesis for every new tonne of lithium supply.

Prices had recovered sharply from their early-2025 lows by 2026 while remaining far below the 2022 peak, investment by lithium-focused companies fell by roughly 40 percent in 2025, and announced projects narrowed the projected 2035 supply shortfall;^{5, 6} under the IEA’s 2026 stated-policy assumptions, EV battery deployment still approaches 3 TWh by 2030⁷ and lithium records the strongest growth among the major critical minerals through 2040, yet future tightness now depends more visibly on project execution, recycling, battery chemistry, policy, capital availability, and the actual pace of demand.⁸

The analytical problem is therefore no longer a directional call on lithium prices: demand growth is real, the supply response is real, and the economic value of new production increasingly depends on where each tonne can sit on the cost curve through an ordinary commodity cycle.

Subscribe now

Seen in that context, the question surrounding new lithium supply becomes more specific: resource size establishes the geological opportunity, while commercial value emerges through the production system that converts that resource into a qualified chemical product at a competitive cost and with sufficient operating reliability; for brine projects, flow, impurities, water, reagents, energy, recovery, selective-medium durability, downstream conversion, availability, and capital structure all accumulate into the economics of the final tonne.^{9, 10}

Direct Lithium Extraction sits directly inside that problem: rather than eliminating the surrounding flowsheet, DLE moves selective lithium recovery into engineered separation systems operating on much shorter residence times and with a different physical footprint from conventional evaporation-based routes, so commercial significance depends on whether the complete system can sustain real-brine performance, utilities, recycle, downstream purification, and product quality over industrial operating periods.^{11, 12, 13}

That shift in emphasis also changes the most useful way to examine the startups now approaching commercial scale, because the relevant question is no longer confined to what their extraction systems can demonstrate in isolation, but extends to the industrial conditions under which that performance has to be reproduced, integrated, and sustained. In this analysis, those conditions are followed backward from the requirements of a full-scale plant through the technical, manufacturing, operating, and commercial layers that support it.

The three startups examined here provide useful views of that transition because each has moved beyond laboratory experimentation while approaching industrial scale through a materially different allocation of technology, manufacturing, project responsibility, and capital.

One is building around a proprietary ion-exchange medium, internal media production, and an expanding role in project integration; another is combining selective extraction with membranes, refining, manufacturing, and direct participation in resource development; and a third is advancing a thermally regenerated liquid-extraction process alongside established engineering and project partners.

They are selected because each places the principal scale-up burden in a different part of the production system, exposing three distinct ways in which field evidence has to be converted into manufacturing consistency, engineering scope, capital commitment, contractual interfaces, and sustained plant operation before demonstrated DLE performance becomes commercial production.

Read more

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

A $1 billion equity round is difficult to miss.

But the more revealing number may be the one beside it: $200 million in debt.

And for advanced nuclear, that distinction matters.

The sector has been financed largely around technical possibility, regulatory progress, first-of-a-kind engineering, and the enormous amount of equity required to survive the journey between them.

Credit asks a different question.

It asks what part of that journey has become defined enough to finance against.

The financing follows a reactor milestone under a U.S. Department of Energy program, meaning the conversation is beginning to move beyond whether the underlying technology can work and toward whether reactors can be manufactured, delivered, contracted, and financed repeatedly.

This week, The Big Idea explores whether advanced nuclear is reaching the point where different risks can be financed differently—and what that could mean for the role of debt across companies, supply chains, and future deployment.

Subscribe now

Beyond this, the edition tracks more than 50 developments across AI infrastructure, optical connectivity, specialized inference, automotive systems, navigation, sodium-ion storage, robotics, defence autonomy, industrial biotechnology, offshore energy, carbon removal, fusion, geothermal, mining, batteries, and advanced manufacturing.

Several numbers stand out: a $5.5 billion valuation in optical interconnects, a $3.3 billion valuation ahead of first customer silicon, a $10 billion compute agreement, more than 1,000 parts manufactured aboard a moving naval vessel, and a rare-earth production roadmap moving from kilograms toward industrial tonnage.

The macro layer then maps eight policy and infrastructure shifts shaping where future capacity can be built and financed.

The edition closes with 10 startups to watch across AI infrastructure, autonomy, energy, advanced manufacturing, and critical materials.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Can Nuclear Build a Credit Layer?

A new financing package suggests advanced reactors may be approaching the point where technology risk, manufacturing risk, and deployment risk no longer need to sit on the same balance sheet.

Advanced nuclear has spent years trying to prove that reactors can become smaller, faster to build, and more repeatable. Its next constraint may be less technical: whether the market can finance each layer of industrial risk with the form of capital best suited to carry it.

Equity can fund possibility. Credit needs a defined claim, a credible repayment path, and some understanding of what remains valuable when schedules slip. The expansion of credit alongside a much larger equity round therefore says something different from another large venture financing.

Recently, a financing package pairing $1 billion in equity with a $200 million credit facility placed that distinction at the center of the advanced-nuclear market.

Read more

Dear friends,

Last week marked two milestones for Deep Tech Catalyst: its 130th episode, and exactly one thousand days since I released the very first one.

1,000 days in which I have consistently worked toward one specific vision: building the playbook I wish I’d had when I started my journey into Deep Tech.

I thought this could be the right time to share a few reflections with you on its origins, where it stands today and where it is going next.

Subscribe now

One of the strongest motivations that pushed me to start came from a feeling of disorientation I had back then and still remember very clearly today.

I was coming from the scientific world and stepping into the intersection of technology, entrepreneurship, and capital for the first time, with a technical background and zero experience in business or finance.

I remember sharing that feeling with many of my scientist friends who, like me, had an entrepreneurial mindset, whenever the conversation moved from applied science to topics such as fundraising, ROI, margins, or distribution.

It was clear that venture capital needed compelling, investable plans before being deployed.

But what kind of plan makes sense when it takes five to seven years to reach the market? What roadmap could justify a budget for the first 36 months? And later on? And then, how do you answer the question, “Can this be scaled up?” if the prototype is literally still on paper?

That wasn’t clear at all. At least for me.

After all, who had ever heard of those questions? Weren’t years spent learning chemistry, physics, or physiology enough?

I was starting to fear they were not…

At the same time, everyone around me kept repeating: “Build, measure, learn.” Basically, build as if it were software. However, hardware seemed to be a bit different. It seemed to follow different rules, different timelines, different models.

As a result, I failed many times.

However, not being able to understand where the gaps were in bringing a diagnostic device, an advanced material, or even a quantum technology to market (and, consequently, into the real world) was simply unacceptable.

Bridging that gap was so important then, and it still is today, because I firmly believe that the ultimate goal of science is serving the people by helping them solve their most critical problems.

What could be more important than that?

One day, after weeks of sharing ideas with investors, entrepreneurs, and industry experts, I realized three things:

1. The SaaS model didn’t fit deep tech.

2. There was no real way to efficiently build VC-backable Deep Tech companies.

3. There was a clear need to decode this path and, if possible, build a common language capable of making the exchange of ideas and value more fluid.

It became clear that bridging the gap between science and capital was not only important but necessary for many reasons.

Necessary to make the world a better place. And I don’t mean that figuratively. I mean helping secure energy, water, food, medicine, materials, and everything that allows humankind to build a civilized society.

Moreover, it was necessary for me, because I wanted to find a way to see science and technology flow more effectively from the laboratory bench into the real world. And that was incredibly fascinating.

So, basically, I decided to start by building the playbook I wish I’d had when I started.

A reference point that would have allowed builders around the world to find it a little less damn difficult to understand how to build an investable business plan.

In late 2023, despite some talented pioneers having already defined its concepts and macro-dynamics, the term Deep Tech often sounded like something absolutely mysterious, capital-intensive, and inexplicably difficult to reconcile with a plausible return on investment.

So, in a world dominated by software, engaging with the intersection of hardware and capital would have been rather new territory.

As I always say, in Deep Tech, success is a complex scenario at the intersection of various multidisciplinary factors, such as technologies, talent, capital, B2B industry demand, procurement, go-to-market strategy, and policy.

If The Scenarionist was born to provide an independent map of evolving scenarios and their dynamics, Deep Tech Catalyst was meant to catalyze action among builders and backers through a common narrative, helping them identify opportunities, reach those scenarios successfully, and ultimately tackle some of the greatest challenges facing the world and humankind.

So where to start? With three simple steps.

1. Start from the financial pull.

2. Produce it independently.

3. Keep it constantly evolving.

I assumed that every investor reviews between 150 and 200 pitch decks per year. So, by interviewing investors, I could provide a broad perspective on what makes a Deep Tech company attractive to investors and scalable.

Moreover, the show would also include exited founders and industry experts who had participated in successful projects to provide another perspective that obviously could not be missing: the B2B commercial perspective.

So, I decided to explore the “investability interface” between science and capital in the way I considered as efficient as possible, as fast as possible, and also as broad as possible in terms of background, geography, industry, and company maturity/investment focus.

In one question:

Or even:

That is where I started.

The result was surprisingly interesting, but it required an enormous amount of work and, above all, overcoming some significant barriers.

• First of all, I had never done a podcast before.

• Second, the only times I had ever spoken into a microphone were small presentations in scientific settings (and a few birthday parties).

• Third, there were a significant number of linguistic and terminology barriers to overcome.

Moreover, this project required fairly intense multidisciplinary study of the subject to design top-notch episode themes and, of course, epic T-shirts and hoodies.

However, it had to be done, so I did it. And the result was an incredible experience.

I learned an incredible amount of things, and I did it in public, sharing it with the community every week.

Moreover, I had the honor of meeting some of the brightest people in the international Deep Tech landscape: some because they joined the podcast as guests, some as mentors, some as friends to share ideas with.

And some as all of those things together.

One thing I found interesting was that, initially, I thought our community would be made up predominantly of founders. Instead, more and more investors and industry leaders joined.

A community that wanted to stay tuned into the state of the art at the intersection of company building and capital stack design.

After one thousand days, Deep Tech Catalyst has delivered more than 2,500 minutes of educational content, as I navigated across a wide range of macro-verticals, including energy, materials, specialty chemicals, semiconductors, quantum, space, defense, robotics, healthcare, mining, water, and agrifood.

After a while, it became clear that not all projects align with VC return expectations, since deep tech is truly heterogeneous, and almost every case has to be analyzed on its own as an independent scenario.

However, I discovered that, across industry, technology, geography, and also team composition conditions, there are similar patterns that I started identifying, mapping, and sharing on the platform.

Moreover, some recurring themes on our podcast led to the development of thematic areas at the intersection of capital and Deep Tech.

If I had to distill them even further to capture the common language DTC has built, I would frame it around five core areas:

All of them are continuously shaped by execution, team, milestones, customers, supply chains, regulation, and strategic partnerships.

Preparing for the next chapter…

At this point, after everything I have built over the past 1,000 days, I wanted to take stock of the journey.

So I returned, as I often do, to a specific quote from Reid Hoffman:

And fortunately, there has been plenty of embarrassment here, beyond my strong Italian accent. Quite a lot of it. There has also been a bit of anxiety, but also a lot of enthusiasm and fun.

However, times are changing and I feel I have to evolve. Deep Tech today is not the same Deep Tech it was in 2023.

In a short period of time, I have had the pleasure of observing the change not only in the markets, but also in the very conception of Deep Tech as a discipline.

It has become less and less a term associated with complexity and unfeasibility, and little by little it is beginning to resonate again as a word, as a trend, as an ecosystem, as a way of doing things, increasingly understood and widespread among people in the industry.

Many more projects are being born. And solid ones.

Many more incubators, accelerators, and funds are moving in this direction, because it is inevitable that, with Deep Tech being so important, there would be support from everything that represents the market’s response to the enormously important unmet needs that Deep Tech can address.

However, these changes also bring with them the need to adapt the format, so I have decided to pause the production of Deep Tech Catalyst for a few weeks, with new objectives in mind, different, more modern, and better suited to changing times.

In the meantime, The Scenarionist will continue at full speed, exploring scenarios, case studies, and analyses every week.

There are a lot of new things coming soon, including new ways for builders and backers to catalyze opportunities, collaborate, succeed together, and create a better future.

But I wanted to take a moment to thank you from the bottom of my heart for joining me, week after week, on this wonderful journey.

1,000 days have passed, 130 episodes, but this is only the beginning of something that has yet to come out.

And I am more determined than ever to discover it together with you all.

Warmly,

Nicola

Disclaimer

All materials published by The Scenarionist are provided solely for informational and educational purposes. They do not constitute financial, legal, or tax advice, nor do they constitute a recommendation, offer, or solicitation to buy or sell securities, crypto assets, or any other assets. The Scenarionist strives to provide accurate information but does not guarantee that its materials are complete, current, or free from errors, and assumes no liability for any reliance placed on them. Any external partnership or sponsored content will be clearly identified as “Brought to you by.” Unless otherwise disclosed, the companies and technologies mentioned do not sponsor The Scenarionist and are not included in its investment holdings. Artificial intelligence tools may be used for operational support and content refinement.

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

2026 has been a year in which the financing architecture around Deep Tech has become harder to ignore.

Companies are raising larger rounds, entering public markets, expanding manufacturing capacity, and building infrastructure whose capital requirements extend well beyond the traditional venture cycle. Yet one part of the capital stack remains notably thin.

Less than one-tenth of the disclosed funding raised by European Deep Tech companies this year came in the form of debt.

It is a European data point, but the question it raises is global: what conditions would allow debt to play a larger role in Deep Tech?

This week, The Big Idea asks whether the answer lies in stronger guarantees and more specialized lenders—or whether debt simply belongs to fewer stages, assets, and business models than the market often assumes.

Subscribe now

Beyond this, the edition maps more than 40 milestones across AI interconnects, battery storage, autonomous logistics, post-quantum security, robotics, quantum validation, advanced nuclear systems, fusion magnets, photonic networks, critical-mineral recovery, synthetic fuels, and distributed AI infrastructure.

The macro layer follows with the forces shaping market access: Europe’s Cyber Resilience Act, industrial access to research infrastructure, the UK’s proposed £1 billion pension-backed scale-up fund, Australia–Japan quantum cooperation, the Philippines’ emerging launch market, asset-level economics for industrial AI, and new regulatory pathways for gene-edited crops.

Finally, the edition then closes with 10 startups selected for how closely they map to this week’s most consequential frontier shifts.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Deep Tech’s Debt Desert

9.2%. Less than one-tenth of the disclosed funding raised by European Deep Tech companies in 2026 came in the form of debt.

The number comes from Tech.eu’s latest 2026 funding snapshot. It is a European data point, but the question it raises is global: what conditions would allow debt to play a larger role in Deep Tech? Do we need better indicators, stronger guarantees, more specialized lenders—or should we accept that debt applies only to certain stages, assets and business models?¹

For those immersed in Deep Tech, these questions have become a recurring part of the conversation. They surface when founders consider dilution, when investors build capital plans for industrial scale, and when policymakers look for ways to mobilize private capital. This week’s data offers a useful occasion to examine them more closely.

One possible response is to reach for a “new menu”:

Read more

Subscribe now

This week, I sat down with Tobias Egle, Senior Associate at Fine Structure Ventures. We explored how advanced materials are becoming increasingly critical to photonics, why foundry compatibility and manufacturing risk matter as much as laboratory performance, how photonics startups can shorten lengthy development cycles, and which customer and foundry commitments provide credible early validation.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

Materials Are Reshaping Photonics

Photonics is becoming a critical part of the infrastructure required to move increasingly large volumes of data. Within that stack, three areas stood out throughout our conversation: lasers, modulators, and packaging.

Each depends on materials, but a compelling solution does not begin with the material itself. It begins with the performance constraint the material can remove.

For instance, a modulator converts an electrical signal into an optical one. Conceptually, the function is simple. Strategically, it influences two of the most important metrics in optical interconnects: energy per bit and bandwidth density.

Energy per bit measures how much energy is required to encode information into the optical domain. Bandwidth density reflects how much information can be pushed through the optical pipeline.

This is where innovation can become commercially meaningful. A new material creates value when it enables better performance at a point in the stack where the industry is actively seeking faster and more energy-efficient solutions.

The material is not the end product, it is what allows the component to address a critical system-level constraint.

Performance, however, is only one side of the challenge. The material must also be turned into something that a semiconductor fab can integrate into its process and manufacture at high yield.

Raw Materials, Supply Chains, and Markets

Photonics bottlenecks do not begin only at the device or fabrication level. In some cases, they extend all the way back to the availability of the underlying materials.

Indium phosphide

Indium phosphide is a key active material for producing lasers, which are expected to become increasingly important as industry roadmaps move toward co-packaged optics.

The challenge is that indium is generally obtained as a co-mined material rather than through a standalone supply chain. This means that the bottleneck can extend beyond wafer production to the availability of the raw material itself.

This illustrates a broader principle in deep tech: a system-level bottleneck can originate far upstream. Even when the need for more devices is clear, growth may still depend on access to specialized materials and reliable sources of supply.

Beyond data centers

AI data centers are a major market for photonics, but they are not the only environment in which new materials and components can create value.

Biosensing presents a different set of requirements. Raw performance may be less important than the ability to integrate photonic components into a broader biological or sensing application.

Quantum technologies can also provide an early proving market.

These applications may require extremely low optical loss because every lost photon can affect performance. Volumes may remain limited, but customers can be willing to pay for components that meet unusually demanding specifications.

Defense can serve a similar role for technologies based on specialized processes. This sector may be more willing to support the transition from laboratory production to an early manufacturing process.

From Lab to Manufacturable Component

A working laboratory device is only the beginning of the company-building process. For a photonics startup, the real challenge is turning that device into a component that can be manufactured at high yield, qualified within a process, and supported by a clear data sheet, proven reliability, and a viable supply chain.

First, a startup should not focus on selling an “innovative material.”

Customers buy a laser, modulator, or other component with a defined data sheet, proven reliability, and a supply chain capable of supporting production.

The material may create the underlying performance advantage, but the component is the commercial product.

This distinction shapes the entire business plan.

The fabless path

A startup will generally aim to follow a fabless model, except in the rarest cases, when it secures the resources required to build a specialized fabrication facility.

However, fabless does not make process transfer a simple external expense. Transferring a process to a foundry is a multi-year engineering program that should be treated as a central part of the company’s roadmap.

The journey often begins in a university research facility, where the first device is fabricated.

The next step is usually a multi-project wafer run at a commercial foundry, where components from several customers are pooled to reduce the startup’s tape-out cost.

This gives the startup its first chip from a commercial foundry, which it can measure and iterate on while improving foundry compatibility.

Iteration speed becomes a strategic advantage

The difficulty lies in the time between iterations.

A startup may wait six to nine months, or longer, between tape-out and the return of fabricated material. Once the device is measured, the team must revise the design and begin the cycle again.

These long cycles can cause a startup to get stuck, particularly because it is unlikely to be a priority customer for the foundry.

One way to reduce the turnaround time is to parallelize development.

The company may tape out at two different foundries or organize parts of the engineering process so that several activities advance at the same time.

The goal is to shorten the turnaround time between tape-out, receiving the fabricated material, measurement, and redesign.

For a photonics startup, the ability to shorten this manufacturing feedback loop can therefore become a strategic advantage.

Subscribe now

The Foundry Needs a Credible Volume Story

A foundry does not evaluate a startup only on the quality of its technology. It also evaluates process and contamination risk, future wafer volume, and the opportunity cost of committing engineering hours and tool time.

From the foundry’s perspective, the central question is whether the startup can eventually translate into meaningful wafer volume.

The foundry ultimately wants to understand whether the process could translate into wafers per month, how many, and whether it could eventually use a significant share of the line’s capacity.

This is why relationships and industry credibility matter. A professor, advisor, or early executive with foundry experience can help the startup gain attention and become a higher priority for the foundry.

New materials can create contamination risks

A new material can threaten the integrity of an existing process. For instance, a metal in one layer may diffuse into an adjacent layer and degrade the device’s performance.

Materials such as gold, lithium, or other metals that could contaminate a CMOS-adjacent line may trigger a detailed contamination review.

Where a contamination risk is identified, the foundry may require quarantine tooling, separate wafer carriers, and strict handling protocols.

In other words, for a foundry, the concern is that a contaminating element could compromise the broader manufacturing process and affect the performance of the resulting chips.

This is why few foundries may be willing to accept the risk of introducing such materials into a CMOS-adjacent line.

The right partner may not be a leading foundry

A tier-2 or tier-3 foundry may be more willing to experiment, particularly if it can dedicate an older line or separate equipment to the process.

Such a foundry may accept the risk in the hope of commercializing a technology that could become an important future opportunity.

Even then, the commercial logic must be strong.

Every hour spent supporting an experimental process is time that could have been allocated elsewhere.

The startup must therefore present a credible path toward meaningful wafer volume that can justify the foundry’s engineering hours, tool time, and process risk.

Building an Investable Roadmap Before Full-Scale Production

In photonics, development cycles can last many years—potentially around five years before a qualified and yielding process reaches production. During that time, competing technologies continue to improve.

A startup therefore needs more than a marginal advantage. A useful rule of thumb is the potential to deliver at least a tenfold improvement in a metric that matters to the customer.

The relevant metric will depend on the specific use case. Ideally, the improvement should be even greater.

The technology must still be compelling by the time the process is qualified, yielding consistently, and ready for production.

Early validation must involve real commitment

A letter of intent may indicate interest, but on its own it provides limited evidence of demand. Stronger validation appears when a customer or foundry commits resources.

A paid evaluation or pilot agreement is more credible because it involves payment or the dedication of engineering hours and fab runtime. A development agreement with defined milestones, clear success criteria, and a well-scoped statement of work is stronger still.

A strategic investment from a customer can reinforce that signal further.

The common element is a customer with the skin in the game. The most credible early agreements show that the partner is willing to commit money, engineering hours, fab runtime, or a conditional purchase commitment.

For investors, this demonstrates more than market interest. It can show that the technology is beginning to connect technical progress, manufacturing execution, and a credible path toward commercial adoption.

Disclaimer

All materials published by The Scenarionist are provided solely for informational and educational purposes. They do not constitute financial, legal, or tax advice, nor do they constitute a recommendation, offer, or solicitation to buy or sell securities, crypto assets, or any other assets. The Scenarionist strives to provide accurate information but does not guarantee that its materials are complete, current, or free from errors, and assumes no liability for any reliance placed on them. Any external partnership or sponsored content will be clearly identified as “Brought to you by.” Unless otherwise disclosed, the companies and technologies mentioned do not sponsor The Scenarionist and are not included in its investment holdings. Artificial intelligence tools may be used for operational support and content refinement.

Subscribe now

Deep Tech companies are usually described through the milestones still ahead: qualification, demonstration, certification, industrialisation, factory build-out, or recurring deployment. This perspective can overlook an equally revealing stage, when the technology has already become valuable to a specific counterparty even though the wider business has yet to reach its intended volume, utilisation, or delivery cadence.

At that point, the company may already have something meaningful to offer. A buyer can purchase a component that fits inside an existing system, material for a formal qualification process, engineering work around a full-scale demonstrator, access to a specialised asset, scientific capacity applied to a defined programme, manufacturing know-how, delivered industrial performance, or a claim on output that will become available after new capacity has been built.

The distinction matters because these early forms of demand carry consequences beyond the first inflow of cash. They influence which work must be completed before payment, who finances the remaining development, where technical and operating responsibility sits, and how much flexibility the company retains as it grows. Two announcements carrying similar headline values can therefore represent markedly different positions and execution burdens.

This analysis examines ten models with case studies spanning space, aerospace, advanced materials, industrial heat, quantum computing, biotechnology, carbon removal, and low-carbon manufacturing.

Subscribe now

Together, the cases show how early demand can strengthen qualification, production discipline, technical evidence, infrastructure utilisation, technology-transfer capability, financing visibility, and confidence in future delivery. It can also concentrate the organisation around bespoke requirements, long-duration obligations, capital-intensive assets, or relationships that become difficult to unwind.

The purpose is to understand how value begins to form before scale and how these early choices influence the company developing around the technology. A final comparison brings the ten models onto common ground, examining what is being purchased, when cash may enter, what capabilities may accumulate, and which risks remain as the company moves towards repeatable delivery.

From here onward, the full analysis is reserved for Premium Members.

1. Early Contracts as Operating Decisions

Read more

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

2026 has been a year in which AI infrastructure commitments have expanded into extraordinary physical quantities: gigawatts of capacity, billions of dollars of hardware, and data-center campuses built around the future requirements of frontier models.

This week, AMD agreed to provide the systems behind up to two gigawatts of Anthropic’s future infrastructure, alongside an investment that could reach $5 billion.

The scale is notable. But the structure behind it invites a broader reflection on the way capital, customers, suppliers, and infrastructure commitments are beginning to reinforce one another. That reflection matters today because such commitments are increasingly influencing valuations and investment decisions across semiconductors, cloud platforms, data centers, power systems, and private capital.

Subscribe now

This week’s Big Idea examines how those layers are becoming connected—and what that may reveal about the signals the market is using to measure the AI buildout.

Beyond this, the edition maps more than 80 milestones across semiconductor architecture, industrial robotics, autonomous transport, AI security, photonic networks, battery storage, advanced nuclear, fusion engineering, textile recycling, bio-based chemicals, orbital infrastructure, industrial water, and critical-material processing.

The macro layer follows with the forces reshaping the market: defence-material traceability, resilient clean-tech procurement, civil nuclear market access, faster autonomous-systems acquisition, public AI infrastructure, space regulation, synthetic-content transparency, science funding, and industrial-policy coordination.

The edition then closes with 10 startups selected for how closely they map to this week’s most relevant frontier shifts, from autonomous logistics and multi-domain defence software to AI materials discovery, perovskite solar, optical switching, and room-temperature quantum memory, followed by 10 Data to Track to sharpen situational awareness before these signals become obvious.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

The AI Compute Loop Eats Its Own Tail

Investors may be treating financially dependent AI commitments as independent evidence of demand.

AMD will supply the systems behind up to two gigawatts of Anthropic’s AI infrastructure while investing as much as $5 billion in the customer and reportedly exploring support for some of its future data-center leases.¹–³

If supplier capital helps finance the buyer whose orders then drive the supplier’s growth, some of the industry’s most impressive demand signals may be considerably less independent than they appear.

Read more

Subscribe now

This week, I sat down with Vincent Berk, Partner at Apprentis Ventures. We explored how AI is broadening the software-builder base and reshaping cybersecurity, why domain expertise is becoming increasingly important as software moves into specialized industries, what creates a defensible moat when products are easier to reproduce, and how founders should think about team composition, technical validation, enterprise sales cycles, and focused go-to-market execution.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

AI Is Expanding Both the Cybersecurity Surface and the Builder Base

Artificial intelligence is changing cybersecurity in two opposite directions.

1. On one side, it is making software easier to build. Founders, operators, and industry experts can now create applications faster, even when they do not come from traditional software backgrounds. This is especially relevant in industrial and highly specialized markets, where the most valuable opportunities are often understood by people with deep domain expertise rather than by generalist software teams.

2. On the other side, the rapid expansion of software is increasing the number of systems that must be protected. A company builds software, someone finds a vulnerability, and another company develops a way to detect or mitigate it. AI does not change that fundamental dynamic, but it simply accelerates it.

More software means more vulnerabilities

AI can support high-quality software development when it is used inside disciplined engineering processes. But companies are also producing more code under tighter timelines, with growing pressure to launch and iterate quickly.

The issue is not simply whether AI-generated code is better or worse. The deeper problem is volume. More software means more applications, integrations, permissions, and potential failure modes.

The concern is therefore not only the quality of individual applications, but the expansion of the software base. When tighter development timelines contribute to lower software quality, the security risk increases further.

Authentication failures, poorly designed permission systems, and unexpected software behavior can all become entry points. As software becomes more embedded in industrial operations, vehicles, infrastructure, and physical systems, those weaknesses can affect more than data. They can affect how the underlying business operates.

AI accelerates vulnerability discovery

AI is also making it faster to search for weaknesses.

A security researcher can inspect software, test common failure modes, and look for vulnerabilities. AI can perform many of the same activities at much greater speed. It may not always be better than an experienced expert, but it can repeat analytical processes across far more code and far more systems.

This increases the overall capacity to search for software weaknesses and accelerates the ongoing cycle of software development, vulnerability discovery, and the creation of new detection and mitigation methods.

The result is a much larger effective workforce capable of searching for weaknesses. Vulnerabilities that might previously have remained undiscovered for longer can now be identified more quickly.

This makes cybersecurity increasingly relevant across a growing range of industries, particularly as AI becomes more deeply integrated with the physical world.

More specifically, as software becomes easier to create, it is moving deeper into sectors that were previously less digitized. Security depends on understanding how a system is supposed to work, what should be permitted, and what kinds of failure are possible.

In increasingly specialized software markets, that knowledge often comes from direct industry experience. Therefore, the people who understand those sectors may be particularly well positioned to become cybersecurity founders.

Domain Expertise Is Becoming the New Entry Point

Cybersecurity is often treated as a field reserved for founders with deep technical backgrounds. That assumption is becoming less reliable.

As AI lowers the barriers to building software, founders with strong industry knowledge can move closer to product creation. They may not have spent years writing code, but they understand how a specific process, system, or market is supposed to work.

That knowledge matters because cybersecurity begins with failure modes.

A security problem emerges when a system behaves in a way that should not be possible. An authentication process fails. A permission is too broad. A piece of software accepts an unexpected input. A device communicates with something it should not trust.

To identify those risks, a founder needs more than general technical ability. The founder needs to understand the operating model behind the system.

Security is embedded in the domain

Every cybersecurity solution is built around an implicit model of normal behavior.

To protect a system, a company must know which actions should be allowed, which should be restricted, and what kinds of behavior indicate that something has gone wrong.

In highly specialized industries, those answers are often not obvious to outsiders.

A founder who understands a particular industrial process may be better positioned to identify domain-specific failure modes that outsiders may not immediately recognize.

The same principle can apply across traditional industries, industrial operations, IoT systems, and connected vehicles.

As software becomes more specific to individual industries, security is likely to become more specific as well.

The market may therefore produce more point solutions: focused products designed to address a particular failure mode, workflow, or operational environment rather than broad platforms attempting to secure everything.

This specialization reflects how software itself is evolving. Bespoke applications can create equally specific security requirements and failure modes.

The New Moats in Cybersecurity

As software becomes easier to build, defensibility becomes harder to prove.

A product may be technically impressive and still become easier to reproduce as software development becomes more accessible. General technical knowledge alone is therefore less likely to provide a durable advantage.

A useful way to frame defensibility is through three elements: process, problem, and people.

Process, problem, people

A process moat comes from knowing how to discover, detect, operate, or solve something in a way that is not obvious from the outside.

A problem moat comes from understanding how a complex system behaves in the real world. This often requires substantial experience with the specific conditions, constraints, and failure modes of that environment.

A people moat comes from trusted relationships. Early-stage products are rarely perfect. Founders need customers who are willing to test incomplete solutions, provide feedback, and tolerate mistakes. Access to those first users can create the initial validation that a startup needs before the market is ready to trust it more broadly.

This is why relationships within a specific industry can be strategically important. They can provide the early access and trust required to test a product, make mistakes, and validate the initial solution.

Real-world performance is a critical test

One of the first differentiators to examine is also one of the simplest: whether the product works outside the lab.

Many cybersecurity solutions are convincing in concept. They combine data sources, identify a suspicious pattern, and appear capable of detecting an attacker.

The difficulty is whether the same approach works inside a real enterprise.

Networks are larger, more fragmented, and more complex than test environments.

A system may perform well on one computer or a small dataset and fail when it must analyze vast numbers of devices, identities, data streams, and connections.

As in the physical world, scalability is therefore part of defensibility.

Team Composition and AI’s Limitations

The quality of a cybersecurity startup depends not only on its technology, but also on the composition and dynamics of its founding team.

At the pre-seed stage, the ideal team is usually small. The conversation suggested a founding team of more than one person and, ideally, no more than three.

However, what matters most is not the number of people. It is how well they work together.

The founders should know one another well enough to challenge assumptions, expose weaknesses, and disagree without damaging the relationship.

They should be able to call out one another’s flaws while maintaining enough focus and trust to work effectively together.

Strong founding teams often combine 3 capabilities.

1. The first is imagination. A team can benefit from someone capable of seeing possibilities beyond the current product: new applications, larger markets, and different ways the technology could evolve.

2. That energy must be balanced by execution. Another member of the founding team needs the executive discipline to reduce the number of possibilities and return the company to its immediate priorities. The team may see ten interesting directions, but it still needs to decide which customer, problem, and market it is building for today.

3. The third capability is domain expertise. At least one founder should understand the industry and the problem deeply. This has become more important as software development has become easier. When the ability to build is less scarce, the quality of the problem insight becomes more valuable.

At its core, technical capability alone is not enough. The team also needs to understand the market it is entering and the specific problem domain in which the product must operate.

AI is powerful, but not universally effective

The same discipline is required when evaluating the role of AI inside the product.

New technologies often appear capable of solving almost any problem before their limitations become visible. AI is still passing through a similar phase.

It is remarkably effective at generating language, knowledge, and code. It can compress a large amount of information into a relatively small model and produce useful output across many different subject areas.

Cybersecurity, however, often depends on a different type of reasoning.

A security system may need to connect a series of failed login attempts with a later successful login, determine whether those events are related, and compare them with behavior across other machines, users, and applications.

That becomes difficult when the events occur far apart in time or across different data sources.

An experiment may work across three controlled data streams. A large enterprise may contain hundreds of thousands of data streams generated by computer systems, identities, and pieces of software.

The technical challenge is not merely identifying an unusual event. It is correlating disjointed events across time and across systems at enterprise scale.

This is where many AI-based cybersecurity products face their real test.

A compelling demonstration may rely on a novel prompt that works in a controlled laboratory environment. That does not prove the solution can identify the same patterns inside a complex enterprise.

Investors therefore tend to look beyond the “AI layer”. The central question is whether the founders have solved the core problems of correlation and scale within their particular problem domain.

AI can help founders build software faster and can assist in developing algorithms to address these problems. But it does not remove the need to understand the underlying architecture of the problem.

The more credible cybersecurity startups are therefore likely to combine an effective founding team, deep knowledge of the problem domain, and a realistic understanding of what AI can and cannot do.

Subscribe now

Cybersecurity Go-to-Market Strategy

A technically strong cybersecurity product does not become a scalable company simply because the problem is real.

Adoption is slow because customers rarely deploy a new security solution across the entire organization from the beginning. They start with a limited test.

They want to know whether the product solves the problem, whether it works inside their environment, and whether it continues to perform as usage expands.

Only after those questions are answered does the customer buy more.

This is the logic of land and expand.

A startup enters through one site or another limited deployment, then grows after the product has demonstrated that it works and scales.

Founders need to reflect this reality in their financial planning. Even when the product works and scales, it may take 12 to 18 months before the company begins to see meaningful traction.

The ideal customer profile must be narrow

Cybersecurity founders can easily begin with an unfocused assumption: everyone has the problem, so everyone can be a customer.

That may be broadly true, but it is not focused enough to support an effective go-to-market strategy.

A startup needs a precise ideal customer profile and a clear understanding of the individual buyer inside that organization.

The company should know who owns the problem, how it will earn that person’s trust, and how it can get in front of the same buyer repeatedly.

Without that focus, the message becomes too broad to attract anyone in particular.

Customer size also matters. Selling to a small company can take as long as selling to a large enterprise because both customers will want to test whether the solution works and scales before expanding its use.

The difference appears after the first deployment.

A small customer may have limited room to expand. A large enterprise can roll the same product out across more locations and at a much greater scale. The initial sales effort may be similar, but the economic opportunity within the larger customer can be significantly greater.

This makes customer selection part of the business model.

Partners cannot replace founder-led sales

Another common assumption is that a reseller or systems integrator will solve distribution.

The logic appears attractive. The partner already has customers, so it should be able to add the startup’s product to its portfolio and generate sales.

In practice, large integrators may already carry vast numbers of products. They have little incentive to invest time in understanding an unknown solution unless they believe it can generate meaningful volume.

The vendors most likely to receive that attention are usually the vendors that are already large.

An early-stage startup should therefore retain control of its initial go-to-market rather than assuming that a partner will create demand and sell the product on its behalf.

Partners may play a role, but founders should not expect them to create the company’s initial traction.

Precision marketing over broad visibility

The same discipline applies to marketing.

Broad brand marketing is a weak substitute for a focused strategy built around a defined ideal customer profile and buyer persona. Trade shows, advertising, public relations, and sales development are only tools.

Their value depends on whether they reach the person responsible for the problem.

The more useful question is not what percentage of the budget should go to marketing. It is whether the company is “shotgunning or sniping.”

A broad outreach approach begins with a generic budget allocation. The company buys ads, attends large conferences, and hires sales representatives without knowing precisely where its target buyer spends time.

A precision approach begins with the target buyer.

The founder identifies the person responsible for a specific security issue inside the type of company the startup wants to serve.

The next step is to understand where that person reads industry news, exchanges ideas, attends meetings, and builds trust with peers.

Sometimes the right channel will be a focused conference. It may instead be a smaller industry group, a standards meeting, a technical community, or a specialized publication.

The most valuable environments are often those in which buyers validate products among themselves.

A cybersecurity company may begin winning customers inside one vertical not simply because the product appears suited to that industry, but because the security professionals within that vertical know one another, meet regularly, and create internal validation.

Once that pattern becomes visible, the company can focus on the places where those conversations happen.

The marketing budget should follow that insight.

Travel, events, articles, advertising, and sales development should be selected according to their ability to reach a known buyer.

If the company cannot identify that buyer with precision, it has probably not completed enough market research.

A realistic financial and go-to-market model should therefore account for long sales cycles, focused customer selection, direct access to the buyer, and the gradual economics of land and expand.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape.

Deep Tech Briefing is the weekly independent intelligence for decision-makers operating across the Industrial Frontier.

Each edition turns fragmented signals across frontier sectors into market context, allocation implications, strategic watchpoints, and the clarity required to compound knowledge into capability.

Subscribe now

A seed-stage company is usually defined by what it has yet to build: its team, technology, customers, and operating system.

This week, a robotics company entered the market from a very different starting point.

The headline is difficult to ignore: a unicorn at seed.

Yet the more important question is not how much capital the company raised. It is what had already happened before the financing was announced.

Subscribe now

This week, The Big Idea examines a less visible route through which deep tech companies can enter the market—and what it could mean for founders, investors, corporations, and the institutions where technical capability is built.

Beyond this, the edition follows more than 50 developments across defence, robotics, quantum, energy, semiconductors, industrial biotechnology, advanced manufacturing, and AI infrastructure.

Capital continued to move toward physical systems, but the most valuable evidence came from execution: major contracts, operating deployments, customer orders, regulatory progress, production scale-ups, and strategic infrastructure commitments.

The macro layer maps the policies and institutions influencing the next stage of industrial growth, including new support structures for energy, semiconductors, defence technology, critical minerals, and compute infrastructure.

The edition closes with 10 companies to watch across robotics, energy, quantum, defence, advanced materials, and industrial biotechnology, followed by 10 Key Data points highlighting the scale, production capacity, and operating evidence behind this week’s most consequential developments.

Enjoy the read!

Deep Tech Briefing is available for Premium Members

By joining The Scenarionist Premium, you also unlock access to exit analyses, case studies, rumors, exclusive lessons, scalability frameworks, and hard-earned insights from more than 100 Deep Tech founders and investors.

Subscribe now

The New Route to a Seed Unicorn

This week, a robotics company came out of stealth with $300 million in seed capital and a $1.1 billion valuation. The headline is striking: a unicorn at seed. Yet the more consequential signal may lie in how the company reached that starting point—and whether its formation suggests that the next generation of deep-tech companies could emerge through a different institutional route.

Read more

Subscribe now

This week, I sat down with Sabrina Sasaki, Partner at Monozukuri Ventures. We explored how scientific founders can move from laboratory validation to industrial adoption, why milestones must reflect the physical complexity of the product, and how Japan’s manufacturing ecosystem can support international commercialization through pilots, co-development, and strategic partnerships.

Key takeaways from the episode:

BEYOND THE CONVERSATION — STRATEGIC INSIGHTS FROM THE EPISODE

Physical AI Becomes Investable When Hardware Meets Industrial Reality

Physical AI is often discussed as the next frontier of artificial intelligence: a world in which machines can perceive, interpret, and act within physical environments.

But from an investment perspective, the category cannot be reduced to smarter robots or more capable automation systems. Its real significance lies in the convergence of hardware, software, data, manufacturing, and industrial adoption.

Hardware gives an AI system access to the physical world. Sensors allow it to observe. Actuators allow it to respond. Robotic systems allow it to perform tasks. Yet the hardware alone does not necessarily create a defensible company.

A growing part of the value comes from what happens after the physical interaction has been captured.

Software must interpret the information generated by the system, identify which signals matter, and convert those signals into useful operational decisions. In industrial environments, this may mean supporting predictive or preventive maintenance, or helping operators understand what action should be taken next.

This must shape the logic of venture building.

A hardware startup cannot assume that engineering complexity will automatically translate into commercial differentiation. The technology must be connected to a broader system of data, intelligence, and business value.

At the same time, the company must demonstrate that the technology can perform effectively in real customer environments, not only under controlled laboratory conditions.

That requires more than a working prototype. It requires a team capable of learning from customers, adapting to uncontrolled environments, financing capital-intensive milestones, and choosing partners that reduce rather than increase execution risk.

A Prototype Is Not Yet a Commercial Solution

Deep tech companies often begin in laboratories, research institutions, or highly controlled engineering environments.

That is where the underlying science can be tested. It is where researchers can isolate variables, measure performance, and demonstrate that a technical concept is possible.

This work is necessary. But it can also create an incomplete sense of validation.

Researchers naturally speak with other researchers. They encounter similar technical problems, use a common vocabulary, and operate within environments that share many of the same assumptions.

A new component or system may therefore receive strong interest from other universities, laboratories, engineers, or designers building specialized prototypes. Those early users may understand the technical achievement immediately.

The difficulty is that validation inside this community does not necessarily prove that a large industrial customer has a commercially meaningful problem.

A technology can be valuable to another research team and still have a limited path toward broader commercial application.

Leaving the controlled environment

The transition becomes particularly difficult when the technology involves robotics, physics, chemistry, or industrial automation.

Inside a laboratory, the founders can control the environment. They can determine how the system is positioned, which variables are introduced, and how each experiment is conducted.

A customer environment operates differently.

Conditions may change, unexpected variables may affect the system, and the founders may no longer be able to control the experiment in the same way.

Field robotics makes this gap especially visible. A machine that performs perfectly under controlled conditions may behave very differently when the environment around it changes continuously.

The same applies to materials and chemical systems. Physical and chemical processes may behave differently when they are moved from a controlled laboratory into a real customer environment.

This is why a perfect laboratory prototype is not necessarily close to a commercial product.

One purpose of early customer engagement is not simply to receive confirmation that the technology is impressive. It is to discover which assumptions fail once the system enters the real world.

Building toward the commercial problem

At the earliest stages, the startup may not be able to build the version that a large corporation will ultimately purchase.

That final product may require too much capital, too much time, or too much engineering work. The level of maturity and performance expected by the customer may be impossible for a young company to achieve immediately.

The startup still needs to demonstrate progress.

An effective approach is to identify shorter milestones that move the technology closer to the commercial need.

A second version of the prototype may improve one critical part of the system. Each milestone can test and improve part of the solution while showing customers that the technology is advancing.

These steps do not constitute full commercial traction. They show that the company is learning in the right direction.

Investors need evidence that the founders understand the gap between what currently exists and what a customer will eventually pay for.

They also need to see that the team can break that gap into milestones that are technically credible and appropriate to the scale and resources required by the product.

Learning from customer feedback

One of the hardest transitions for scientific founders is learning how to respond to negative feedback.

Positive feedback is comfortable because it confirms the original hypothesis. It reinforces the value of the science and supports the idea that the founders are building the right solution.

Negative feedback can be especially useful.

It may reveal that the customer does not experience the problem in the way the founders expected. It may also challenge the original assumptions about the application or show that the solution needs to adapt to what the market requires.

The point is to apply the same willingness to test assumptions to the market.

Founders can treat customer interactions as another source of evidence. They can test assumptions, examine contradictions, and adjust the product when reality does not support the original design.

Subscribe now

Milestones Must Match the Scale of the Product

It is difficult to discuss comparable roadmaps for physical AI or industrial deep tech. The capital required to develop a sensor is not comparable to the capital needed to build a large robotic infrastructure system.

An actuator can sometimes be prototyped using existing components. A new tunneling platform, energy system, or aerospace product may require substantial resources before a credible minimum viable product can even exist.

The investment stage alone does not determine what progress should look like. The scale and technical nature of the product are equally important.

At pre-seed, the team is the main evidence

The initial market thesis may change. The use case may evolve. The founders may discover that their differentiation is different from what they originally believed. Customer conversations can redirect the product, and early experiments can expose limitations that reshape the entire company.

For that reason, pre-seed investment is often primarily an assessment of the founders and the relationship among them.

The important questions concern how the team works under pressure, how responsibilities are divided, and how the founders respond when experienced people challenge their assumptions.

Accelerators, mentors, and advisors can be useful partly because they create this pressure. They force founders to explain why the company is different, where the value comes from, and what evidence supports their claims.

The purpose is not simply to improve the pitch. It is to observe whether the founders can process criticism without becoming defensive or losing conviction.

Early-stage company building requires both.

A team must be confident enough to pursue an uncertain technical vision and flexible enough to change when the evidence demands it.

Seed stage

By seed stage, investors expect more than a scientific concept, but what counts as meaningful progress varies widely.

For a component company, it may be possible to build multiple prototypes, test them with customers, and improve the design relatively quickly.

Off-the-shelf parts may reduce development time, and the startup may be able to reach a functional demonstration with limited resources.

For a company developing a more complex infrastructure technology, each prototype may be expensive. The company may need specialized equipment, large testing environments, or physical installations before it can prove that the system works.

The first few units can cost disproportionately more than later production.

The relevant question is whether the company has defined milestones that reflect the scale of what it is building.

A credible plan must show what can be learned from each unit of capital. It must explain which technical risks will be retired, which assumptions will be tested, and how each development stage brings the company closer to a product that customers can adopt.

Japan’s Industrial Ecosystem and Opportunities for Startups

Japan remains an attractive environment for industrial and manufacturing technologies because much of its competitive strength exists below the level most founders initially see.

International entrepreneurs usually recognize the country through its large consumer brands. They know the automotive companies, electronics groups, and industrial names that became symbols of innovation during previous decades.

But many of those companies built their products through networks of highly specialized suppliers.

Small and medium-sized manufacturers developed deep expertise in specific materials, components, processes, and production techniques. In many regions, clusters formed around a narrow category of industrial capability.

Their specialized knowledge and manufacturing capabilities may be valuable resources for startups to access.

Industrial customers rarely behave like early adopters

A corporation will not normally accept technical risk simply because the startup is ambitious. Industrial companies are generally unwilling to take risks on behalf of startups, particularly when evaluating solutions for business-critical environments.

The required quality level is therefore much higher.

This is particularly important when working with large manufacturers, whose reputation has been built around rigorous quality control.

A startup cannot approach these companies with the expectation that they will tolerate an unreliable product while it develops.

Before pursuing a serious industrial relationship, founders must understand the quality standards against which their technology will be evaluated.

Corporate interest does not necessarily indicate commercial progress

Many large companies are still learning how to work with startups. Their internal processes may move slowly. Decision-making authority may be unclear.

The innovation team may be interested in the technology without controlling the budget or business unit required for adoption.

On the other hand, an early-stage company has limited time and limited capital.

A large corporate conversation can appear strategically important because the brand is recognizable and the potential contract seems significant. The startup may invest months in meetings, technical discussions, custom requests, and internal presentations.

Yet the relationship may not move the company toward its real milestones.

Therefore, before achieving product-market fit, identifying meaningful use cases, securing sufficient financing, or developing a stable product, extensive corporate engagement could become a distraction.

This does not mean founders should ignore large companies until they are fully mature. They can begin building the relationship earlier.

They can educate potential partners, share product developments, communicate new use cases, and keep relevant stakeholders informed about fundraising and technical progress.

The distinction is between nurturing a relationship and depending on it.

Conversely, once the startup has reached a greater level of maturity, the conversation with a corporate partner can become more productive.

A company with a stronger product, credible customer use cases, and a clearer market position is better prepared to engage with an industrial corporation.

Before reaching that point, founders can keep potential partners informed about the company’s progress, applications, customers, and fundraising activity without relying on them to drive the startup’s immediate milestones.

This is especially important because some corporations may not fully understand how to work with companies that have not yet reached product-market fit.

The objective is to pursue industrial partnerships at a stage when they can genuinely support the company’s development.

The Right Industrial Partnership Depends on What the Company Needs

There is no single model for working in industrial B2B markets. The right structure depends on the technology, the resources required to build it, and the strategic importance of the market.

For advanced materials and semiconductor companies, collaboration with an established manufacturer can be highly valuable. These startups may need equipment, laboratories, process knowledge, and production infrastructure that would be unrealistic to recreate independently.

A larger industrial partner may already possess the machinery and technical teams required to test the technology at a different scale.

In those situations, co-development can create interesting opportunities.

The startup contributes a new technical approach. The manufacturer contributes facilities, production knowledge, and the ability to translate the concept into a repeatable process.

The challenge is negotiating partnerships that allow both sides to test and develop the relationship effectively.

One option is to begin with a limited agreement that allows both parties to test the relationship. The collaboration may run for a defined period to understand whether the technical work and working dynamics are effective.

This can reduce uncertainty.

Both sides can use this period to understand whether the collaboration works and whether it can support further technical development.

The purpose of the initial relationship can be to create evidence.

It should demonstrate that the companies can build something together and provide a basis for deciding whether a deeper partnership makes sense.

International commercialization considerations

Opening an operation in another country consumes management attention and financial resources. It also requires the company to commit significant time and organizational effort to that market.

A local presence becomes rational when the company’s strategy is genuinely tied to a specific ecosystem.

This may be the case when partners are essential to manufacturing, when the country is a core customer market, or when local grants and industrial partnerships can materially accelerate the business. In other situations, a local representative or commercial partner may be sufficient.

In other words, the operating model should follow the company’s real dependency.

For instance, an established industrial partner can provide the infrastructure and capabilities needed to manufacture the product while helping control manufacturing risk according to established standards.

This can unlock resources that would be difficult for the startup to obtain independently. Moreover, public funding may become more accessible when the company has a credible plan to manufacture locally with a respected partner.

Furthermore, a manufacturer that already serves customers in other Asian industrial economies may become a route into those ecosystems.

But the value of these options depends on how well the partnership model aligns with the company’s strategy and technological needs.

A semiconductor component, an advanced material, or a complete robotic system require different resources and different forms of collaboration.

In conclusion, the company must begin with a clear understanding of what it needs from the partner, what it can develop independently, and how the relationship will help it reach the next stage of commercial development.

Disclaimer

Please be aware: the information provided in this publication is for educational purposes only and should not be construed as financial or legal advice or a solicitation to buy or sell any assets or to make any financial decisions. Moreover, this content does not constitute legal or regulatory advice. Nothing contained herein constitutes an offer to sell, or a solicitation of an offer to buy, any securities or investment products, nor should it be construed as such. Furthermore, we want to emphasize that the views and opinions expressed by guests on The Scenarionist do not necessarily reflect the opinions or positions of our platform. Each guest contributes their unique viewpoint, and these opinions are solely their own. We remain committed to providing an inclusive and diverse environment for discussion, encouraging a variety of opinions and ideas. It is essential to consult directly with a qualified legal or financial professional to navigate the landscape.