Full-scope engagement
AI due diligence leads; technical and product diligence run underneath it and report into the same investment view. One team, one evidence base, one ranked set of findings.
The frame
Measured performance against the task the buyer is actually paying for.
Claims traced to evidence: evals, logs, code, not demo footage.
What is owned versus rented from a foundation model provider.
Behaviour under load, latency budgets, failure modes.
Inference economics and gross margin at realistic usage.
Whether AI features drive adoption, retention and price.
Data moat, switching costs, model-provider commoditisation.
Team depth, evaluation discipline, AI governance maturity.
Process
Sequenced to your process and timeline, with interim findings raised as soon as they are material rather than held to a final read-out.
Capture the investment thesis and every material AI claim, then convert them into testable propositions and an evidence request list.
Code, pipelines, evals, traces, telemetry, provider contracts and invoices. Management and engineering sessions run against specific claims.
Technical findings read against product usage and against the economics. Most material findings only appear where the three views intersect.
Ranked findings, thesis matrix, risk register and a sequenced post-close agenda.
Engagement tiers
Three levels of depth, scoped to the access you have and the decision in front of you.
Engagement tier
Pre-LOI, early-stage, competitive processes, or where access is limited.
Covers
Typically 2–4 business days
Request an AI Red-Flag ReviewEngagement tier
Most commonA committed process where the AI claims carry the investment thesis.
Covers
Typically 5–10 business days
Discuss Full AI DiligenceEngagement tier
Complex, regulated or enterprise AI, AI acquisitions and agent platforms.
Potential tracks
Scoped to the transaction
Discuss ScopeWork product
Each track contributes to one integrated investment view.
Each thesis assumption tested against evidence rather than against the management narrative.
| Assumption | Evidence | Finding | Assessment | Investment impact |
|---|---|---|---|---|
| AI accuracy is materially ahead of incumbents | Internal eval set, 3 customer benchmarks, no held-out test set | Advantage holds on two of six task types; eval set overlaps training data | Partially supported | Narrow the win-rate assumption; fund an independent evaluation before close |
| Gross margin expands with scale | 12 months of provider invoices mapped to usage | Inference cost grows faster than seat revenue on agentic workloads | Not supported | Assume margin compression until pricing moves to consumption |
| AI features drive expansion revenue | Product telemetry, 20-account cohort, expansion ledger | Expansion concentrated in accounts using two of five AI features | Supported, narrowly | Prioritise activation of the two proven features in the value creation plan |
Solid cyan = proprietary and owned. Dashed amber = an external dependency the target does not control.
Revenue indexed to 100. Inference cost and residual gross margin as monthly token volume scales.
Monthly tokens processed (x). Margin compression at scale is a pricing-model question, not an infrastructure question.
Defensibility by stack layer. The filled block marks where the layer actually sits; block height rises with strength.
Model
Third-party foundation models, no fine-tune ownership
Data
Labelled outcome data accumulated per customer
Workflow
Embedded in approval chain, replicable in ~2 quarters
Integrations
Six systems of record, each a switching cost
Customer Data
Tenant-scoped history improves retrieval quality
Brand
Category awareness undifferentiated in buyer interviews
Distribution
Two channel partners, concentration in one
Every feature is marketed to 100% of eligible accounts. Bars show the share that used it in a trailing 30-day window.
Bring the thesis, the data room and the timeline. We will tell you what evidence exists, what is missing, and what it means for the deal.
