Key Takeaways

  • Software revenue ramps smoothly, but physical hardware runs near zero for two to five years before spiking straight into tens of millions of dollars upon deployment.
  • Early-stage hardware diligence cannot rely on traditional SaaS growth metrics because prototypes require long cycles of research, component testing, and certification.
  • Team evaluation shifts between rounds, moving from evaluating founder technical ability at pre-seed to assessing the caliber of hires drawn from operators like SpaceX and Anduril by Series B.
  • Letters of intent, demonstration pilots, and strategic balance-sheet checks from prospective enterprise buyers replace billings as the primary signals of commercial demand.
  • Investors evaluate capital efficiency and execution speed through Polovets' 3-Pillar Early-Stage Deep Tech Evaluation Framework.

The Polovets 3-Pillar Early-Stage Deep Tech Evaluation Framework

Software underwriting models break down when applied to physical systems. As David Weisburd observed during the conversation, “You can't just vibe code it and then you have product market fit and then you're scaling. You have to build hardware. Sometimes it's a missile. There is no minimum viable missile.”

Because commercial revenue remains absent during initial rounds, Leo Polovets relies on three specific signals to underwrite early-stage deep tech ventures through their pre-revenue development:

Pillar 1: Team Strength

At pre-seed and seed it's the founders, but then towards Series A and B, it's what kind of team are you recruiting? Is it top folks from SpaceX and Anduril or is it new grads that don't know much about hardware and you're trying to train them? Those are very different team profiles.

Pillar 2: Technical Milestones

Clearly moving at a fast speed towards building something hard. Went from design to components to subsystems to the whole system in about three years which I think is very fast.

Pillar 3: Market Validation

Usually that's LOIs or letters of intent. It might be pilot contracts or demos or future customers that are investing in your company because they're excited about what you're doing and they've seen what you got planned.

When This Works (and When It Doesn't)

This framework applies directly to seed, Series A, and Series B rounds in capital-intensive sectors such as defense, energy, semiconductors, and data center hardware. In these markets, engineering cycles take years to yield an integrated asset. When enterprise or government buyers sign non-binding letters of intent or commit early pilot capital, they establish that a market awaits the hardware once physical assembly and regulatory testing finish.

Where this framework breaks down is late-stage Series B extensions and Series C rounds. At that threshold, technical progress and non-binding letters of intent lose their explanatory power. An LOI carries zero legal obligation to convert into a binding purchase order. If a company transitions from subsystems to a complete assembly but discovers unit economics are underwater, early customer excitement will evaporate. Teams assembled from elite defense primes can solve intricate physics problems while remaining blind to supply chain yields, scrap rates, and factory tooling capital expenditures. At scale, technical progress without contract conversion becomes an expensive science project.

Why It Matters

As artificial intelligence commoditizes pure application software, venture capital is rotating out of traditional SaaS and into physical infrastructure, energy systems, and hardware bottlenecks. This transition forces underwriters to abandon smooth annual recurring revenue multiples in favor of binary technical milestones and high-value, step-function contract ramps. The capital allocators winning in this environment are those structuring financings around physical engineering progress rather than premature commercialization.