AI Due Diligence
A company in your pipeline says AI is its edge. In 3–10 working days I check how much of that holds up: code, data, model costs and team. You get a memo with a recommendation: invest, renegotiate or pass.
Who it’s for: For VC and PE funds, angel investors and M&A teams assessing an AI company, or a company whose results AI could improve or threaten.
AI due diligence doesn’t replace legal or financial review. I assess the technology, the costs and the team.
Every company has an “AI edge” now.
In a pitch deck, AI looks the same at a company that built something hard to copy and at one that wrapped someone else’s model in a nice interface. The difference shows in the code, the data and the API bill: what it costs to serve one customer, what happens when the model vendor raises prices, and whether the team can measure quality. I check that before you sign the term sheet.
What you get
- An assessment of the edge: what is proprietary (data, processes, evaluation) and what is a layer on someone else’s model
- AI economics: cost per customer or task, sensitivity to vendor prices, margin at scale
- Quality and risk: how the company measures model output, what happens on errors, dependence on a single vendor
- Team: whether it can build and run AI, or only integrates ready-made tools
- A memo with a recommendation: invest, renegotiate or pass, with questions for the investment committee
- For the full review: a 100-day post-close plan with the two changes that matter most
How it runs
Ground rules
To keep the assessment independent:
Price
from €5,000net · red-flag review
The starting price covers a red-flag review (3–5 days) of one company and its key AI claims. The full review with a 100-day plan starts at €13,000 net.
- a fixed fee agreed before starting, independent of the deal outcome
- the start date is set around your investment committee
Tell me which company you’re assessing and when the investment committee meets. Within 1–2 business days I’ll reply with a price and a start date.
Or by email: michal@jaskol.ski
As of: November 2026
Before you write.
How is this different from technical due diligence by a large consultancy?
It’s done by someone who builds AI products and has twice taken a company from startup to IPO. I look at what the company’s CTO looks at: code, costs, quality. The memo is a few pages long and ends with a recommendation.
How much access do you need from the company?
For a red-flag review: the data room, a demo and two hours with the CTO. A full review needs repository access (read-only is fine), a data sample and model bills from recent months.
Do you also assess companies that don’t build AI?
Yes, when the question is whether AI threatens their business model or could improve margins. The memo then covers risks and opportunities, and the 100-day plan names processes to automate.
What happens after the deal?
You can hand the 100-day plan to the company’s management. If the company wants, I can deliver it as a Sprint, only after closing and at the company’s request.
What’s next
After the deal, the 100-day plan goes to the company’s management. Delivering the most important change is natural material for an Agentic Workflow Sprint commissioned by the company.