The Matrix for Investors

AI due diligence for venture capital and private equity. Place a deal, or every portfolio company, on the Vector & Value Matrix: where AI creates value, where its risks sit, and which claims have not been checked.

Try the diligence demo Read the case studies

Risk and governance due diligence only. Not investment advice, and never a recommendation to buy, sell, or hold. Read the disclaimer.

What investors get

The matrix answers an investor’s version of the leadership question: where is AI creating value in this company, and where is the risk it brings?

A deal profile

A company placed on the matrix through six investor questions, with its zone, its red flags, and the diligence steps that follow, ready for an investment committee memo.

A portfolio compass

Every holding placed and weighted by position size, showing where the portfolio’s AI exposure leans against the external Matrix Vector.

Red flags

Unverified AI claims, dependence on a single AI vendor, margins exposed to AI costs, and volume outrunning expert review.

Verified claims

Where a company’s value rests on AI performance, ReasonIQ audits test whether its benchmark claims hold up.

The six investor questions

The same two axes as the matrix, asked from an investor’s side of the table. Each is answered on a four-point scale.

Revenue at stake. How much of revenue depends on AI-driven products or processes?
Expert review. Does an expert review consequential AI outputs before they reach customers?
Judgment. How much does the product’s value rest on expert judgment rather than automated output?
Upside. How much does AI improve growth or margins?
Margin exposure. What happens to gross margin if AI costs double or the main AI vendor changes terms?
Regulation and rights. Is AI used in regulated or high-risk decisions, or on data the company does not clearly own?

How AI due diligence works

Engagements run alongside your existing commercial, financial, and legal diligence, and start with a mutual non-disclosure agreement.

Scope
Agree which companies, claims, and data rooms are in scope, and sign the NDA.
Assess
Place each company on the matrix from management answers, documents, and interviews.
Verify
Test the AI claims that matter most to the investment case, with independent reproduction where needed.
Report
Deliver a findings report and investment committee pack: zone, red flags, and the questions to resolve.

Case studies

Two illustrative cases, one for venture capital and one for private equity. The firms and companies are fictional; each portfolio opens in the demo.

Venture capital: Series B diligence

The startup whose accuracy claims nobody had checked

Lantern Ventures (fictional), a venture fund reviewing a Series B round in ClaimPath AI (fictional), which automates insurance claims decisions, alongside six existing holdings.

Open this portfolio in the demo

The situation

ClaimPath’s pitch rested on a headline accuracy figure and fast growth. The partners wanted to know where the company’s AI value and risk really sat, and how the deal compared with the rest of their AI portfolio.

What the matrix showed

ClaimPath sat in Automate, finely balanced between value and risk, and was flagged as volume outrunning expert review: claims decisions that need judgment were only spot-checked.
Its accuracy claim had not been independently verified, it depended on a single AI vendor, and its margin was exposed if AI costs doubled.
Across the fund, five of seven holdings depended on a single AI vendor, and five had unverified AI claims.
The holdings leaned toward machine scale, between Automate and Authorize, only partly aligned with external pressure toward risk carried.

What the investor did next

Asked for an independent audit of ClaimPath’s accuracy claims before completing its own investment process.
Reviewed the AI vendor contract, including pricing terms and change-of-control clauses, and modelled gross margin with AI costs doubled.
Prioritized verification for the health and hiring holdings, where AI decisions carry regulatory risk.
Added the six investor questions to the fund’s standard diligence checklist.

Private equity: portfolio review

The buyout portfolio leaning away from where the risk was building

Harbourline Capital (fictional), a private equity firm with six portfolio companies (fictional) adopting AI in logistics, healthcare, insurance services, consumer goods, and engineering.

Open this portfolio in the demo

The situation

Operating partners had approved AI projects company by company. The investment committee wanted one view of where AI was creating value across the portfolio, and where it was adding risk that value-creation plans did not yet cover.

What the matrix showed

The portfolio leaned toward machine scale, and away from where external pressure was pushing, toward risk carried.
Over a third of the portfolio by size (38%) sat in Authorize, driven by personalised pricing at Vista Retail and claims intake at Northgate Brokers.
Northgate was flagged as volume outrunning expert review, and BrightPath Clinics sat close to the judgment line in clinical triage.
Four of six companies had AI claims that had not been independently verified.

What the investor did next

Asked Vista Retail and Northgate to review data rights, vendor terms, and regulatory exposure as part of their value-creation plans.
Asked BrightPath to keep clinician review on every triage decision and test whether reviewers catch errors.
Added AI governance to each portfolio company’s board agenda, with the matrix re-run every quarter.
Planned independent verification of AI claims ahead of any future exit process, so buyers’ diligence finds no surprises.

Confidentiality

Every engagement begins with a mutual non-disclosure agreement, signed before any information about a fund, a deal, or a company is shared with us.

Information is used only for the engagement it was shared for, and is not disclosed to third parties without written consent.
We do not use information from one engagement in another, or in our research, case studies, or marketing.
On request, we return or delete materials at the end of an engagement, subject to any legal obligation to retain them.
Please do not send deal information through our website contact form; we will arrange a secure channel once the agreement is signed.

Disclaimer

Not investment advice. Not a recommendation to buy, sell, or hold. Our AI due diligence, the Matrix for Investors, the diligence demo, and any report or pack we produce assess AI-related value, risk, and governance only. They do not rate, value, or recommend any company, fund, or security, contain no price targets or investment views, and must not be read as a recommendation to buy, sell, hold, or otherwise deal in any investment.

Not a regulated adviser. Vector & Value AI is not registered as an investment adviser or research analyst with the Securities and Exchange Board of India or any other securities regulator, and does not provide investment advice or research reports within the meaning of those regulations.

Limited scope. Our work covers AI-related matters only. It is not financial, accounting, tax, legal, or regulatory due diligence, and it relies on information provided by the companies and investors involved, which we do not independently audit unless an engagement expressly includes verification.

Your decisions remain yours. Investment decisions are made solely by investors and their advisers, who remain responsible for their own diligence and for compliance with the laws and regulations that apply to them. No particular outcome is guaranteed.

Illustrative case studies and demo. The firms, companies, figures, and actions in the case studies and demo samples are fictional and do not describe any real investment, client, or result. The demo is provided as is, without warranties of any kind, and data entered into it stays in your own browser.

Need AI due diligence on a deal?

We assess a company’s AI claims, dependencies, and risks under a mutual NDA, and report to your investment committee.

Talk to us