Questions to Ask Before You Commission a Technical Review
A checklist for non-technical buyers: who does the work, what access they need, what you receive, and what happens if the result is bad.
No hot takes. No thought leadership theatre. Just what I'm learning while building software and running a consultancy.
What you can check yourself about software an agency or freelancer built, what needs an independent view, and who owns the code, repositories and hosting.
Mike Nash
A checklist for non-technical buyers: who does the work, what access they need, what you receive, and what happens if the result is bad.
Most codebases that feel broken can be fixed. Some can't. How to tell which you have, why rewrites often fail, and what an independent review can tell you.
The questions a good technical due diligence report answers, the red flags it should surface, what a clean result looks like, and how long it takes.
What drives the price of an independent code audit, the published fixed-fee tiers on this site, and what a cheap automated scan does and doesn't cover.
Most people don't understand how Claude's prompt caching works. I didn't either, until I looked at my own token reads and writes and built a tool to fix them.
Most AI-assisted builds have the same security gaps. Authentication looks right until you trace it. Here's what I find when I audit before go-live.
Before you go live with an AI-assisted build, there's a question worth being able to answer honestly. Most people can't. Here's what that costs.
Three clients in a month, each about to go live with an AI-assisted build, all asking the same question: is what they got actually solid?
The practical guide to setting up coordinated AI agent teams that actually deliver.
How I built a skill that captures and preserves learnings from each development session.
How I use Claude Opus as a sounding board for architectural and product decisions.
The gap between prototyping with AI and shipping to production - what it actually takes.
My AI agents were getting unruly - silent failures, no visibility. So I built a real-time control room to fix it.
How I set up a Notion board with autonomous AI agents picking up tickets, doing the work, and updating status.
Deep dive into the context management and workflow setup behind disciplined Claude Code usage.
Exploring disciplined AI-assisted development with Claude Code - building an autonomous workflow that actually stays on track.
Why bad code happens, who's really responsible, and what to do about it.
Reflections on how AI has changed the software development industry and where we're headed.
LLMs, agents, prompt engineering, and production AI systems
Architecture, technical decisions, and building systems that scale
Managing teams, making decisions, and growing as a technical leader
AWS, Azure, infrastructure, and keeping costs under control
Building things people want, shipping fast, and learning from users
Running a technical consultancy, working with clients, and pricing