What Is an Architecture Diagnostic? Process, Duration, and Outcome
An Architecture Diagnostic is a structured assessment of your codebase with a weakness analysis and roadmap — here's how it works in practice, typically completed in under 14 days.
An Architecture Diagnostic is a structured assessment of an existing software codebase: it surfaces weaknesses, makes concrete recommendations for change, and delivers a roadmap for further development — regardless of whether that development happens in-house afterward or with an external partner. It isn't a rewrite, and it isn't a finished piece of software — it's the foundation both of those actually need to build on.
Why diagnose first, instead of building right away?
In an earlier post, a team faced a feature nobody wanted to touch — and made the call to take it apart first, before building anything further. An Architecture Diagnostic makes that same kind of decision systematically, across the entire codebase, before a single line of new code gets written. Without this step, teams either keep building blindly — risking making existing weaknesses worse — or rewrite everything prematurely, even though large parts of the existing solution might be solid.
How it works in practice
The Diagnostic follows four steps: first, a review of the existing codebase and a conversation about the team's biggest pain points. Then the actual analysis, following the same three criteria from the earlier post — complexity, interfaces, testability. The result is a written architecture and roadmap document. It wraps up with a video call where the results are walked through together.
How long does it take, and what does it cost?
Typically under 14 days, depending on scheduling and the complexity of the codebase. The entry price is a flat fee of €3,500.
Who owns the results?
All materials — including the architecture and roadmap document — become your property. That holds regardless of whether you continue working with me afterward, hire a different partner, or implement the plan in-house.
Who is this for?
Mostly for teams that built a product with AI coding tools that now has paying customers, or that's approaching a funding round — the moment a working prototype turns into a responsibility the original structure can no longer carry.