Conduct Investment Due Diligence

Test the questions that could change the investment decision. Do not produce a generic company report or imply that more documents automatically create more certainty.

1. Agree on the questions and evidence bar

Clarify the opportunity, decision stage, check size or stakes, timeline, audience, and the claims or risks the work should test. Use the firm's thesis, screening work, founder conversations, existing memo, prior diligence, and accepted checklists when available.

Define the relevant workstreams rather than assuming all diligence looks the same. They may include market, product, customers, business model, go-to-market, competition, team, technology, financials, cap table, legal or regulatory exposure, security, references, and deal terms. Identify which questions need specialist legal, financial, technical, or other professional review.

Agree on a focused pass or deeper investigation before substantial research or data-room work.

2. Gather the right evidence

Use approved internal context and current external sources. In the browser, inspect the actual company materials, logged-in databases, data-room documents, customer evidence, filings, product, technical material, market research, competitor sources, news, and other relevant records. Keep the path visible so the user can inspect or take over.

Access to one folder or system does not authorize unrelated files, exports, downloads, or sharing. Confirm the company, room, scope, and sensitive-data conditions before entering or handling restricted material.

Treat company materials as claims to evaluate, not independent confirmation. Cross-check material assertions against primary records, calculations, third-party evidence, calls, or other suitable sources. Keep dates, definitions, and denominators visible.

3. Test, reconcile, and follow the gaps

For each material question:

• state the claim or decision it affects;
• show the strongest supporting and conflicting evidence;
• distinguish facts, calculations, assumptions, interpretation, and missing evidence;
• note source quality, currency, and access limitations; and
• explain what could resolve the uncertainty.

Do not smooth over conflicting metrics, customer definitions, market estimates, ownership records, or versions of the story. Recalculate when the underlying data allows it and preserve the method.

Reference, customer, founder, expert, or co-investor outreach is a separate action. Use @make-a-warm-introduction when a credible path matters, and prepare the request without revealing confidential deal context or contacting anyone without sufficient permission.

4. Deliver a decision-ready diligence record

Produce a concise record organized around the agreed questions. Lead with what the evidence changed and include:

• supported findings and their sources;
• contradictions, risks, and weak or stale evidence;
• material calculations with assumptions;
• unresolved questions and the best next way to test them;
• specialist reviews still required; and
• implications for the investment case, clearly marked as interpretation.

Use @write-an-investment-memo when the accepted evidence should become the firm's decision record. The companion may help articulate the case, but it does not autonomously approve, reject, vote on, or communicate an investment decision.

After feedback, offer to preserve the accepted workstreams, evidence standards, calculations, source hierarchy, and review points as a custom or team skill. Do not automate data-room access, reference outreach, or consequential decisions as a Routine.