Ship with AI agents, safely and repeatably.
The Approach
engineering.
- ~26 years shipping production systems
- Runs AI agent fleets in production
- Test-first · least-privilege · adversarial review
Proof-of-work, not hype. For ~26 years I've shipped production software. Today I run fleets of AI coding agents in production under real discipline, a five-role agent team, orchestrator-first, with a human gate at every step.
That discipline is the point: test-first enforced by a QA gate, capability-sliced least-privilege tools, adversarial review that blocks on real findings, and every decision written down as an ADR. A repeatable SDLC, not a pile of prompts.
I teach engineering teams to do the same on their own codebase, and I name the limits: where agents help, where they don't, and when to keep a human in the loop. Every claim here is backed by code you can read.
Ways to work together
Classroom courses on a shared sandbox codebase, private team days, and EU AI Act sessions. Book a call for a scoped quote.
1 day, open enrolment
One seat, one developer, on a shared sandbox codebase the whole room works from. You leave with a working agent team on your own machine.
1 day, private
The same day run for one team, on the sandbox by default or your own repo on request. One team, one shared practice, agreed in a day.
2-day intensive
Everything in the 1-day, plus safe tools (MCP), parallelism, and your team's written "when we use agents, and when we don't" policy.
Embedded advisory
I ride along on a real project for a few weeks, install the practice for real, and leave your team self-sufficient.
Compliant AI for your team
The same gates, ADRs, and audit trail give you the EU AI Act human-oversight story (Art. 14): compliance as a by-product of good engineering.
Get new write-ups by email
Leave your email and I'll email you when a new write-up is published, no spam, nothing else.
Contact
Book a call
Leading an engineering team that's adopting AI agents? Book a scoped call, or just tell me what you're working on.