What do investors check in technical due diligence?
Code quality, architecture, test coverage, security posture, dependency and licence health, scalability, key-person risk, and the team's development process. For AI-assisted codebases, reviewers increasingly ask how generated code was reviewed and tested.
How do I prepare my codebase for due diligence?
Build the evidence picture before they do: current test and coverage state, resolved or documented security findings, healthy dependencies, and a written risk register with owners. Fix what you can; document honestly what you cannot.
Can bad code kill a funding round or acquisition?
Occasionally — but undisclosed problems are the more common killer. Known technical debt with a credible plan is priced in; discovered technical debt triggers repricing, escrow demands, or walked deals.
What red flags do acquirers look for in a codebase?
No meaningful tests, unresolved security vulnerabilities, heavily outdated dependencies, licence contamination, a single developer holding all knowledge, and release history full of incidents and rollbacks.
When should founders start preparing for technical due diligence?
Six months before you need it, minimum — coverage, findings, and process evidence cannot be backfilled in a week. The cheapest option is to keep the evidence continuously current so preparation is a non-event.
Does AI-generated code hurt valuation?
Not by itself. What hurts is AI-generated code without evidence of review and testing, because the acquirer has to price in re-verifying it. Documented review, tests, and findings around generated code neutralise the concern.