How we work
- Understand the brief. Who uses it, what is not working, what a useful result looks like.
- Agree the first delivery. What you will receive, how we will check it, what it costs and what I need from you.
- Build and demonstrate. Try working software against the outcome we agreed. If your needs change, we discuss the impact before continuing.
- Hand over or continue. Documentation and agreed access, then optional support the team can budget for.
AI agents help me research, code and test. I review their work and remain responsible for what I deliver.
For agencies
If you have a client project and need engineering capacity, let's talk. We agree how I'll work with your team, communicate with the client and hand over the finished work. Your client relationships stay yours.
Built it with AI, then got stuck?
A prototype can get you surprisingly far. The next questions are harder: is the data safe, do the important flows work, and can you change it without breaking something else?
- Review what exists. Walk through the app together and identify the problems that prevent real use.
- Agree a repair plan. Keep what works, explain the changes needed, and set a clear scope.
- Finish and verify. Repair the code, test the important journeys, and explain how to maintain it.
Whether you call it vibe coding or an AI-built prototype, you can bring it as it is. We decide what is worth keeping after looking at it.