AI-first engineering
Speed without design, tests, and review is not a capability. I help teams adopt AI-assisted delivery with the same accountability they would expect of any production change.
The work is advisory, training, and hands-on implementation, including legacy environments. Teams learn how to build with AI on real codebases before they scale process around it. That includes machine-readable standards, review loops, and human ownership of what ships.
What you get
- Engineering environment readiness: standards, tests, and machine-readable rules the tools can follow
- Ways of working: accountability, design-before-build, and review loops that fit AI-assisted delivery
- Hands-on coaching with production constraints, not a tooling demo
- Quality gates: automated tests, CI/CD, validation, and human review of generated change
- Organisation-level adoption: connecting emerging capabilities with how engineering actually operates
Evidence from delivery
I have architected and built an AI-enabled B2B SaaS product from the ground up, directing design and implementation while using AI-assisted engineering in the IDE, with reusable rules, automated tests, and pipelines as first-class controls.
On a high-growth marketplace, I directed a search and catalogue platform onto a legacy stack and took it to production in weeks: AI-generated implementation under human architectural direction, without a production incident on go-live, in a language I had not shipped in before.
I have also led generative AI strategy and adoption for a global online marketplace group, translating tooling into engineering practice rather than a slide-deck programme.
I write in public about this work: code review, testing, safety gates, and retaining human technical accountability. Read the insights.