Generic tech DD templates miss what actually kills deals in 2026: AI claims that aren't production-grade, architecture that won't survive 2× traffic, and eng teams that look senior on LinkedIn but can't ship.
This checklist is what I use when doing independent engineering due diligence for VCs and acquirers ($1,250 fixed one-off, one to two weeks).
Architecture & scalability
- Can the current architecture handle 2× and 10× users without a rewrite?
- Monolith vs. services: was the split intentional or accidental?
- What is the actual bottleneck — database, queue, third-party API, or frontend?
Code health signals
- Test coverage: meaningful tests or coverage theater?
- Deployment frequency and rollback story
- Dependency risk (unmaintained packages, license issues)
- Incident patterns in the last 90 days
Team depth
- Bus factor on critical systems
- Seniority mix vs. what the pitch deck claims
- Hiring plan realism (can they actually close the roles?)
AI / ML claims (often skipped in generic DD)
- Production AI vs. demo-ware: evals, fallbacks, cost controls?
- Model routing and token spend — is there a bill shock coming?
- Data pipelines: where does training/inference data actually come from?
Security baseline
- Secrets handling, auth model, dependency vulnerabilities
- Compliance gaps that block enterprise sales
Deliverable investors actually use
Not a 200-page PDF. A traffic-light summary per area, specific findings with evidence, and recommended questions for management.
Full scope and pricing: https://themuneebh.com/technical-due-diligence
I'm Muneeb Hussain — Head of Engineering running production AI systems, fractional advisor for founders on the side.
