Skip to content
MERVENORBook a triage call

Checklist · Taking over a project

Technical due diligence: what investors check in your code, and how to prepare

By MervenorPublished Last reviewed 4 min read

In short

Technical due diligence asks whether the software is owned, safe, understandable and able to grow, and how much depends on one person. Early rounds look lightly; acquisitions and later rounds look hard, often with an outside reviewer. Most of the work is evidence you can collect in advance: who owns what, how the system is built and run, and what it costs.

What reviewers are looking for

A reviewer is trying to answer four questions: does the company own what it is selling, can it be trusted with customer data, can other engineers keep building it, and will it survive growth. They read the code, ask for documents and talk to whoever built it. Clear written answers shorten the process; surprises late in it cost time and sometimes valuation.

Ownership

  • The source code, the domain, the cloud and hosting accounts and the app-store accounts are held by the company, not by a founder's personal account or a former contractor.
  • Everyone who wrote code, employees and contractors, has signed an agreement assigning the work to the company.
  • Open-source licences are listed, and none of them conflicts with how the product is sold.
  • Paid services and API keys are billed to the company.

Security and data

  • Sign-in, roles and access checks are enforced on the server, not only in the interface.
  • Secrets are kept out of the code and the browser, and can be rotated.
  • Known vulnerabilities in dependencies are tracked and fixed on a schedule.
  • You can show where personal data is stored, who can access it, how long it is kept and how it is backed up.
  • Any past security incident is written up with what was done about it.

How the system is built

  • A new engineer can build and run the application from the README on a clean machine.
  • Tests cover the paths that make money or touch data, and run automatically on every change.
  • Deploys are repeatable, with a test environment and a way back.
  • There is a one-page overview of the architecture and of the main decisions behind it.

Running it and growing it

  • Monitoring and error reporting are in place, and someone is alerted when the application fails.
  • Backups exist and a restore has actually been tested.
  • You know what breaks first at ten times today's usage, and roughly what it would take to fix.

Costs

  • The monthly cost of hosting, databases and third-party services is known, including AI usage.
  • You can say how that cost grows with customers or usage, and where the expensive parts are.

People

  • More than one person understands each critical part of the system, or the knowledge is written down.
  • If the product was built by an agency, a freelancer or AI coding tools, say so, and show that someone now understands and maintains it.

How to prepare

  1. Collect the evidence

    Account owners, agreements, the licence list, the architecture overview, the cost breakdown and the backup test, in one folder.

  2. Fix the cheap red flags first

    Accounts in personal names, secrets in the code, missing assignment agreements and untested backups take days, not months.

  3. Get an outside review before the investors do

    A short independent review finds what you cannot see yourself and gives you a written report to share.

  4. Write down what is not fixed yet

    A known issue with a plan is a normal finding; an issue the reviewer finds first is a surprise.

Questions people ask

  1. When does technical due diligence happen?

    Usually after a term sheet or letter of intent, before the deal closes. Early rounds may only ask a few questions; acquisitions go much deeper.

  2. Will an app built with AI coding tools fail due diligence?

    Not because of the tools. It fails if nobody understands the code, the data is exposed, or it cannot be built and deployed reliably.

  3. Do we need a full security audit?

    Not always. A clear picture of the risks and a plan for the important ones usually matters more than a certificate at an early stage.

If this is your application

Auspicia, the launch-ready check, answers the questions technical investors ask about security, scaling, costs and who can run the code, in a written report you can share, in five business days.

For startups

01Start

Tell us about the application

Thirty minutes, no slides, no sales script. You leave with a view on what to check first.

Length
30 minutes
Cost
Free
You leave with
A view on what to check first