Skip to main content
Technical second opinion

Software technical audit before you decide

A software technical audit is an independent review of a platform's code, infrastructure, security and access, done so you can decide on evidence: keep it, refactor it, migrate it, rebuild it or switch vendor.

You need one when a decision is due and nobody in the company can read the code: you inherited the platform from a vendor who left, you have a quote you cannot assess, or the app you generated with AI breaks with real users. We hand you a report with every finding prioritised and a recommendation.

When you need one

Five situations where a second opinion saves the expensive mistake

It is usually commissioned by a company with no CTO or senior engineer on staff that has to decide about software it did not write. Also by founders and investors who need technical due diligence before they sign.

I inherited a platform and do not know what is inside

The vendor who built it is gone, or the freelancer stopped answering. Before reading any code we find out whose name the repository, domain, cloud account and production keys are in. Then we check whether what runs on the server matches the repository, and what it takes to deploy a change without breaking anything.

Repository
whose name is it in?
Domain and DNS
who renews it?
Cloud account
who pays for it?
Production keys
who holds them?
Inventory
access and owners
In your name
your accounts
What we review

Eight areas, from the repository to the contract

The code is one part. The other is everything the software needs to keep running if the team maintaining it changes tomorrow.

Access and ownership

Whose name the repository, domain, cloud account, app stores and payment services are in. Who has access today, including people who no longer work with you.

Architecture and dependencies

How the pieces are split, which frameworks and versions they use, which have lost support and which libraries have gone years without an update. Where the business logic lives.

Security

Secrets in the repository or its history, authentication and permissions, dependencies with known CVEs, and unpatched PHP, Node.js or NGINX versions.

Data and backups

Where the data lives, whether there are backups, when a restore was last tested and whether schema changes are versioned or were made by hand.

Performance and infrastructure

Slow queries, oversized or maxed-out resources, and which part of the cloud bill can be cut without touching the product.

Testing, CI and deployment

What the tests cover, whether deploying is one command or a ritual only one person knows, and how long a rollback takes when something fails.

Documentation

Whether an outsider could run the project locally and deploy it from what is written down. Missing documentation goes into the report as one more finding.

Legal, GDPR and licences

Where personal data is stored and in which country, which providers process it, and whether any third-party licence clashes with how you use it. We flag the technical side; the legal assessment is your lawyer’s.

If something turns up that cannot wait for the report, we tell you as soon as we see it and handle it separately: PHP security patch, Node.js patch, exposed credentials rotation or, if there has already been a breach, hacked website response.

Need to decide what to do with your software?

Get a second opinion →
How we work

From access to decision

The diagnosis runs on read permissions. Handover is a separate phase, and it only starts if you ask for it.

Access
read-only
Analysis
code · infra · data
Findings
impact · risk · effort
Report
decision and roadmap
Handover
optional
Handover only starts if you ask for it, and another team can do it
  1. 01

    Read-only access

    The repository and, where possible, the cloud console and logs. For the diagnosis we do not touch production or ask for write permissions.

  2. 02

    Analysis

    We run the project locally, read the code and configuration, review dependencies and talk to the people who use it and to whoever maintained it, if they can still be reached.

  3. 03

    Prioritised findings

    Each finding gets three ratings: business impact, risk of leaving it as it is, and effort to fix it. That keeps the urgent apart from what can wait.

  4. 04

    Report and decision

    Keep, refactor, migrate or rebuild, with the reasons and a phased roadmap. We present it in a meeting with whoever makes the call.

  5. 05

    Handover, if you want it

    We can start on the high-risk findings, or prepare the handover to your team or to another vendor. The report works the same in all three cases.

What you get

A report to decide with, not to file away

It is written for two readers: the decision-maker, who needs the risk and the recommendation on one page, and the team that will do the work, whether yours, ours or someone else’s.

  • A one-page summary for the decision-maker: what there is, what risk it carries and what we recommend
  • Inventory of access and owners: repository, domain, cloud, app stores and payment services
  • Findings table with impact, risk and effort
  • Recommendation (keep, refactor, migrate or rebuild) with its reasons
  • Phased roadmap, starting with what cannot wait
  • List of the documentation another team would need to carry on

Fictional example: these findings do not belong to any client.

A first read in one minute

Rebuild, refactor or keep?

Four questions we ask in the first conversation. With what you know today they give a first direction; the firm answer comes from reading the code.

If you are unsure about any of them, that already tells you something: it is exactly what the audit answers.

Your case in four questions

Question 1 of 4

Do you have the complete source code and access to production?

Before you sign

What changes when the decision comes with a report

Before the project starts is the cheapest moment to discover that the repository is missing or the database has no backup.

Deciding without an audit
  • You compare quotes that use the same words for different things
  • The new vendor finds out mid-project that access or parts of the code are missing
  • A platform that needed two fixes gets rebuilt, or one that needed rebuilding gets patched
  • Security holes surface when someone exploits them
Deciding with an audit
  • Every proposal is compared against the same list of findings
  • Access and code verified before they change hands
  • The decision to rebuild or refactor has its reasons in writing
  • The urgent items are fixed before the big project starts
What stays in your name

An audit that does not create another dependency

Checking who controls the software would make no sense if the outcome were depending on us.

Code and accounts in your name

If a repository, domain or cloud account turns up in someone else’s name, the first recommendation is to move it to an account of yours. Anything we build afterwards also goes in your accounts.

Documentation anyone can use

The report and any documentation we produce are written for another team to use. They cover how to run the project, how to deploy it and what not to touch without tests.

Confidentiality

We sign the NDA before receiving any access. The report is yours and is not shared with third parties without your permission.

No lock-in

The audit does not tie you to hiring us for development. You can take the report to your own team or to another vendor.

FAQ

Software audit FAQ

Do I have to give you access to the code?
For a full audit, yes: read access to the repository and, if possible, to the cloud console and logs. Read permissions are enough for the diagnosis. If you do not have the code yet, the first part of the work is finding out where it is and how to recover it, and that is already a finding in the report.
What if the current vendor will not cooperate?
We start with what you do control: the domain, the cloud account if it is in your name, the app stores and what is visible in production. From there we prepare the technical list of what is missing and what to request from the vendor in writing. What your contract says about code ownership is for your lawyer to assess; we give them the list.
Do you review apps built with Lovable, Bolt, v0 or Cursor?
Yes. The code they generate gets audited like any other. We look mostly at what these tools do not handle on their own: keys exposed in the browser, database access rules, migrations, backups and monitoring. Sometimes fixing that is enough; other times parts need rewriting before you open the app to more users.
What happens after the report?
You decide. You can fix the urgent items with your own team, take the report to another vendor or ask us to handle it. If we carry on, we start with the high-risk findings. Commissioning the audit does not commit you to hiring us for development.
Do you sign an NDA?
Yes, before we receive any access. The report is yours and we do not share it with third parties, including the current vendor, unless you ask us to.
Is this a security audit or a pentest?
It is not a pentest. Security is one of the eight areas we review (secrets, permissions, dependencies with known vulnerabilities, unpatched versions), but we do not run penetration tests. If the system needs one, the report says so.
Does it work as technical due diligence before investing or acquiring?
Yes. An investor or buyer cares about the same questions: who owns what, how much technical risk there is and how much effort it would take to fix. What changes is the summary, written for whoever makes the investment decision.
How much does it cost and how long does it take?
It depends on the size of the system and the areas to review: an app with one backend is not the same as a platform with several services and integrations. In the free 30-minute consultation we look at the scope and, before starting, tell you what it includes and what it costs.
Let’s start

Tell us what you have and what you need to decide

In the free 30-minute consultation we look at what the software is, what decision you face and what would need reviewing. Then we tell you whether an audit makes sense and what it would include.

Software Technical Audit and Second Opinion

Tell us your challenge. We'll propose a solution.

No commitment. We reply within 24 hours, and within 48 hours you have a proposal with scope, timeline and budget. No fine print.

Book a free call →