Understand the asset before you underwrite it.
Capital rarely loses on the thesis. It loses in diligence — in the gap between the deck's claims and what the technology, the team, and the execution path can actually carry. The flaws that sink a deal are almost never the ones the financial model was built to catch.
Technical Diligence is a first-principles evaluation of deeptech and AI ventures, run for the corporates and funds assessing them — and for the founders preparing to be assessed. It surfaces the defensibility, the readiness, and the failure modes that decide whether a return is real or remediable. Operator-built, from every seat at the table.
Diligence seen from every seat.
I have evaluated technology ventures from inside the rooms where the decision actually gets made — as a screener, as an acquirer, and as a founder being acquired.
The accelerator floor.
Through accelerator and incubator programs, I have screened, mentored, and judged hundreds of startups at the earliest stage — where the question is not “is the model right?” but “is there a there there, and can this team build it?” That volume teaches pattern recognition: which technical claims hold up under a second question, and which dissolve.
The acquirer's side.
Inside corporates, I worked corporate development and innovation from the buyer's side — assessing outside technology for strategic fit, integration cost, and the distance between a working prototype and a manufacturable, shippable product. That is the seat where you learn that the demo is the easy part, and real diligence is everything the demo was designed to skip.
The founder's side of an exit.
And I have sat where the company is the one being examined — having been part of exits to Intel and SoftBank. Living through diligence as the target teaches what the buyer's checklist misses, where founders oversell, and where genuine value hides behind a weak story. Seeing it from both sides is the whole point: I know what the acquirer is looking for and what the target is bracing for.
The rigor underneath it.
All of it traces to Stanford's structured engineering-design tradition — the same first-principles methods that anchor the First-Principles Product track, turned toward evaluation rather than creation. A diligence engagement is product definition run in reverse: instead of building the traceable line from customer need to architecture to failure mode, I pressure-test whether that line exists at all.
The bench that comes with the seat.
Years across accelerators, corporates, and exits don't just build pattern recognition — they build relationships. The most valuable asset I carry out of those rooms is a network of trusted domain experts: the people I screened alongside, the specialists who ran the hard parts of an acquisition, the operators who have actually shipped in regulated and complex domains. No single evaluator should claim depth across every field a complex venture spans, and the credibility of a verdict depends on knowing where your own resolution ends. I lead the engagement, set the structure, and own the synthesis — and where a target reaches into specialized terrain like hardware and manufacturing, biotech and regulated medical devices, or fintech and compliance, I bring in the right expert from that network. The client gets one accountable point of contact and a team assembled to the specific shape of the deal — without overclaiming breadth a review would expose.
Founder-stage screening · Corporate-development assessment · Exits to Intel and SoftBank · Stanford design rigor — and the network built across all of it.
Financial diligence is solved. Technical diligence is where deals quietly go wrong.
Financial and legal review are mature, well-staffed, and widely available. A rigorous read on whether the technology actually works at scale is none of those things.
The financial lens has no resolution here
Investment teams are built to read markets, models, and cap tables — not to ask whether the silicon can be sourced, whether the architecture scales past the demo, or whether the AI system learns or just impressed once. The most expensive flaws live exactly where that lens goes blurry.
Once is not ten thousand times
Most ventures can do the thing once. Diligence is whether they can do it ten thousand times, under load, in production, at a cost that preserves the margin the model assumes. Reproducibility, not capability, is the real question.
Grade the flaw, don't just find it
For a fund or an acquirer, every quarter of unplanned remediation is return erased. The job is to grade each flaw — fatal, or fixable with known cost and known time. A clear go / fix / pass verdict is worth more than a list of concerns.
Three pillars, mapped to the questions a deal team has to answer.
Every engagement moves through the same three pillars — the structured-design toolkit, pointed at evaluation instead of creation.
Defensibility & value-chain analysis.
Where does this venture sit in its value chain, what must it be best at, and does that position compound or erode? Using Customer Value Chain Analysis, I map every stakeholder and what flows between them — money, data, influence — and locate the moat that holds under load versus the one that's a feature in disguise.
Technical & AI-native readiness assessment.
How far is the technology from production reality? For AI ventures specifically: does the system retain feedback, adapt to context, and improve with use — or is it a pilot that plateaus? I assess the architecture, the data position, the evals that prove the system works, and the honest distance from prototype to scale.
Failure-mode & execution-risk review.
What breaks, how likely, how severe, how detectable — and what does fixing it cost in dollars and quarters? Applying Failure Mode & Effects Analysis to the whole system, including the manufacturing, ops, and integration risks that surface only after the deal closes, when remediation is most expensive.
Both sides of the deal.
An outside read you can underwrite
A hard-hitting evaluation of a target's technology, defensibility, and execution risk — delivered on the deal timeline, with a go / fix / pass verdict and the flaws graded by cost and time to remediate.
Find the flaws on your terms
The same evaluation run before the investor's, so the gaps get surfaced and framed by you — not discovered across the table. Walk into diligence already knowing what they'll find and what you'll say.
Find the flaws before they find you.
Bring the target, the thesis, and the timeline. I'll build the diligence that tells you whether the return is real, remediable, or a pass.