AI Technical Diligence
An investor-grade read on a target’s AI claims, before you commit capital.
PE and VC investors, and corp-dev teams, evaluating an AI-related acquisition or investment.
A target company’s AI story sounds good in the data room, but you need to know if it holds up — what’s real and what’s fragile. You also need to know what it will cost to keep running. Vendor and management claims aren’t independent evidence, and generic tech due diligence misses AI-specific risk.
- First days Read the data room
We read the target’s AI claims against the real system and the real data behind it. We flag every claim we can’t verify.
- Mid-engagement Score technical debt and lock-in
We assess the target’s technical debt and its vendor and model lock-in exposure. We score what it will cost to keep running.
- Final delivery Deliver the committee-ready call
You get a written report and a go or no-go recommendation, timed to your deal calendar.
- Claims vs. reality
- We compare the pitch deck and the data room against the system we can actually see running.
- Data rights and lineage
- We check whether the target holds the rights and the lineage to keep using its training and inference data.
- Technical debt
- We score the shortcuts already in the architecture and what it costs to unwind each one.
- Vendor and model lock-in
- We check how hard, and how costly, it would be to switch model or infrastructure vendor.
- Team and continuity
- We check whether the people who built the system are staying, and what breaks if they don’t.
- Security and compliance posture
- We check access controls and data-handling practice against what the deal thesis assumes.
- Technical due diligence on the target’s AI systems and claims
- Data and infrastructure audit scoped to the deal thesis
- Risk register with named owners and severity
- Readiness and technical-debt scoring across the stack
- Assessment of vendor and model lock-in exposure
Everything below transfers to you at handover, no retainer, no lock-in.
- A written diligence report suitable for an investment committee
- A risk register with severity and mitigations
- A go / no-go recommendation
- Technical debt and AI-claims findings
- A signed, deal-ready deliverable
- Not a build or an implementation review. We assess the target; we don’t integrate with it.
- Not a substitute for legal or financial diligence. We cover the technical and data risk only.
- Not a rubber stamp. If the claims don’t hold up, the report says so.
- Skip this if the target hasn’t opened its data room yet. Come back once diligence access is granted.
- Skip this if you need a build-readiness check on your own team, not on a target. The Ship Audit is the fit.
- Skip this if you need an ongoing technical advisor through close. This is a fixed-scope report, not a seat at the table.
Time-boxed to your deal calendar, typically 1–2 weeks
Premium fee, confirmed after first contact
The diligence report is licensed for the requesting investor’s internal use and for sharing with counsel and co-investors. Standard NDA and engagement terms apply, scoped to the deal timeline.
You own every line.#
Explicit handover terms, not fine print — this is what transfers to you when we're done.
- All source code transfers to you at handover.
- Infrastructure-as-code for the whole system transfers to you.
- The eval suiteAn eval suite is a versioned set of test cases that measures whether outputs are good. and tests transfer to you.
- Credentials and accounts we provision transfer to you.
- Documentation and runbooks transfer to you.
- No retainer required to keep it running.
- No kill-switch, no vendor lock-in — operate and extend it without us.
Questions, answered.#
The questions we get asked most, answered plainly: no hedging, no marketing copy.
01 Who is this for?
PE and VC investors and corp-dev teams evaluating an AI-related acquisition or investment, who need an independent technical read before the deal closes.
Link to this answer: Who is this for?02 Can you do investor-side technical due diligence?
Yes. We assess a target company’s AI claims, data practices, and technical debt on behalf of investors, and deliver a written report suitable for an investment committee.
Link to this answer: Can you do investor-side technical due diligence?03 How fast can you turn this around?
We time-box the engagement to your deal calendar, typically one to two weeks depending on scope and target access.
Link to this answer: How fast can you turn this around?04 Does the target’s data leave their environment?
No. We assess in place wherever possible, under the access controls and NDAs already governing the deal.
Link to this answer: Does the target’s data leave their environment?05 Do you talk to the target’s team directly?
Only with your approval and theirs. Some engagements run on data-room artifacts alone. Others need a short technical interview. We scope that with you upfront.
Link to this answer: Do you talk to the target’s team directly?06 What if the finding changes the deal?
That’s the point of running this before you sign, not after. A finding that moves price or terms is worth more here than in month three of ownership.
Link to this answer: What if the finding changes the deal?