‍A Working Definition

Deep tech companies turn breakthrough science or engineering into commercially scalable products. The breakthrough is not a feature or a business model twist. It sits at the core of why the product performs better, costs less, or enables something that previously wasn’t possible.

In practice, deep tech often shares a few traits:

• The work begins with technical uncertainty, not just market uncertainty.
• Progress requires validation, iteration, and engineering rigor before it looks like “a product.”
• Commercialization can involve scale-up, manufacturing, qualification, or regulatory pathways.
• The moat compounds through IP, deep domain know-how, and customer-backed evidence earned in real operating conditions.

‍

What Makes Deep Tech Different

A simple way to differentiate deep tech from is to ask where the value is created.

If value depends primarily on application-layer software, distribution, or workflow packaging, you may have a great business, but it is not necessarily deep tech.

If value depends on novel compute architectures, new materials, new biological capabilities, or systems that must perform reliably in the physical world, you are often in deep tech territory.

Three differences show up early:

• Risks begin as technical. The first questions are “Will it work?” and “Will it keep working?”
• Time is spent proving, not pitching. Progress is measured in tests, pilots, and repeatability.
• Capital is sequenced differently. Up-front R&D may be relatively higher to reach product-market fit – BUT, software application companies can be equally capital-intensive in later growth stages once customer acquisition and competition-driven spend kick in.

‍

Sectors Where Deep Tech Increasingly Shows Up

Deep tech spans many sectors, but it tends to cluster where engineering, science, or infrastructure enables new opportunities for increases in performance or net-new capabilities altogether:

• Semiconductors and Compute Systems: performance-per-watt, bandwidth, latency, reliability, cost.
• AI Infrastructure: training and inference efficiency, systems reliability, hardware-aware software.
• Robotics and Autonomy: sensing, perception, control, edge intelligence, safety-critical operation.
• Advanced Materials and Manufacturing: better strength, yield, weight, cost curves, or sustainability.
• Biotech Disruption: diagnostics, imaging, automation, and discovery that move capability closer to the point of need.
• Energy and Industrial Decarbonization: technologies where reliability and economics must work at scale.

This list is not an exhaustive taxonomy by any means. It’s a reminder that “deep tech” describes an innovation and commercialization pathway as much as a category.

‍

From Uncertainty to Evidence

Deep tech companies are built by converting breakthroughs into businesses. The path is different for every company, but the overall arc is consistent.

• Proof: Demonstrate the core effect or capability in controlled conditions.
• Prototype: Show repeatability and define failure modes.
• Pilot: Prove it works in a real environment with real constraints for real customers.
• Scale-Up: Systematically transition to manufacturable, serviceable, and reliable.
• Expansion: Turn early wins into a repeatable deployment model and durable unit economics.

‍

The Deep Tech Investor Playbook: Underwrite A De-Risking Roadmap

Deep tech diligence is not a longer version of SaaS diligence. It is a different discipline. The best deep tech investors fund milestones and not narratives, by asking questions like:

• What is the single biggest technical unknown right now?
• What evidence would eliminate it?
• What does it cost, in time and dollars, to produce that evidence?
• What new risks appear once this one is solved?
• What customers believe in this product today?

‍

Treat IP As Strategy, Not Paperwork

Patents matter, but deep tech defensibility is usually a portfolio:

• IP plus process know-how
• Engineering discipline and systems integration
• Data and validation evidence (where relevant)
• Manufacturing, qualification, and ecosystem positioning that competitors can’t shortcut

‍

Make Manufacturability A First-Class Risk

For hardware-anchored deep tech, “works” and “works at scale” are different products. Investors should pressure-test:

• yield and test strategy
• supply chain dependencies
• reliability under operating conditions
• certification and qualification timelines

‍

Diligence The Deep Tech Adoption Path, Not Just The Market Size

Deep tech is often won by the first credible deployment, not the biggest TAM slide. Key questions:

• Who is the first buyer, and what do they risk by adopting?
• What validation do they require (standards, testing, third-party proof)?
• What is the integration burden, and who owns it?
• What is the ROI case that makes early friction rational?

‍

Bring Deep Tech “Intellectual Capital,” Not Only Financial Capital

In deep tech, financial capital is necessary but rarely sufficient on its own. Nascent companies benefit from investors who can contribute operating knowledge from repeated scaling experiences: how to sequence product decisions, avoid predictable commercialization traps, and reduce the volatility that slows execution.

This is especially true when founders come from academic or deeply technical backgrounds and are building their first company at production scale.

‍

The Deep Tech Founder Playbook: Lead With The Claim, Then The Proof

Deep tech founders do not need to oversimplify. Make the breakthrough clear, state the core technical claim in one sentence, and then show how you validate it:

• What you tested
• What passed and what failed
• What you learned
• What’s next, and what it will cost

‍

Choose a Deep Tech Beachhead Before You Build a Platform

Platforms are real. Markets still buy outcomes. Pick a first use case where:

• The pain is acute and measurable
• The buyer is reachable
• The deployment path is realistic
• The early win builds evidence that expands the platform later

‍

Design For The Real World Early

Deep tech companies earn credibility through reliability:

• Repeatability beats one-off demos
• Edge cases matter
• Serviceability and integration are part of the product
• Manufacturing readiness is a milestone, not an afterthought

‍

Deep Tech Funding: Match Capital To Milestones& Pick Investors Who Can Help You Ship

• The cleanest raises are tied to a clear “next proof” that reduces risk. 
• A strong plan also mixes capital types where appropriate: non-dilutive programs, strategic partnerships, and later, financing structures that fit scale-up.
• Deep tech is harder to evaluate than software. It is also harder to support. Founders should favor investors with domain pattern recognition and the operational judgment to help navigate the unavoidable complexity.

‍

Common Deep Tech Myths

Myth: Deep Tech Always Requires More Capital.
Deep tech may require more capital earlier to prove and engineer the product, but software often becomes equally capital-intensive later through customer acquisition and competition-driven spend. The difference is sequencing.

Myth: Deep Tech Can’t Produce Venture-Scale Returns.
Deep tech can take longer to mature, but once it reaches go-to-market readiness, growth can accelerate quickly. Products that deliver step-change performance or cost improvements can become standards.

Myth: Deep Tech Is Just Hardware.
Many deep tech companies are hardware-anchored, but deep tech also includes foundational infrastructure, bio platforms, cryptography, and tightly coupled hardware-software systems where engineering depth is the moat.

Myth: Deep Tech Is The Same As A Moonshot.Not all deep tech is a science project. The most investable deep tech pairs real technical advantage with economically meaningful problems and clear applications.

‍

Why Celesta Included A Point Of View

This playbook is intentionally high-level, but it reflects one practical lesson: deep tech is won by teams who pair technical ambition with disciplined execution.

Celesta is a specialist deep tech investor founded in 2013. Our team is intentionally built around a diverse and complementary set of business building experience, not just capital allocation. We emphasize technical diligence, IP depth, manufacturability, and the practical steps required to move from breakthrough technology to scaled global business. We also view intellectual capital as a differentiator in deep tech, because the experience of being a seasoned operator can reduce the avoidable missteps a startup often experiences and keep progress on schedule.

‍

Closing Thought

For founders, deep tech is the work of turning proof into product. For investors, it’s the work of funding the milestones that make that transformation inevitable. For the two of us together, it’s a partnership that changes the world.

‍