Consulting & mentoring

Pragmatic AI for engineering teams.

I help technology leaders and their teams to put AI to work across the engineering process, make their operations agentic, and ship quality software, backed by proven engineering practices.

Engineers trained in disciplined, test-first AI-assisted development
50+

Engineers trained in disciplined, test-first AI-assisted development

Years coaching engineering teams on TDD, BDD, and continuous delivery
17+

Years coaching engineering teams on TDD, BDD, and continuous delivery

Books published on testing and AI-assisted development
3

Books published on testing and AI-assisted development

Published papers, including peer-reviewed research on agentic refactoring
2

Published papers, including peer-reviewed research on agentic refactoring

Ways to work together

Four ways I can help.

Pick whichever matches where your team is right now.

01

AI advisory

For CXOs, founders, and heads of engineering

Where AI actually fits in your business, what to build and what to buy, and which workflows are worth making agentic first. I still build and ship AI products myself, so the advice comes from inside the work rather than from a deck.

  • An unbiased read on vendors and where your money is going
  • Which parts of your operations are worth making agentic first
  • Advice from someone still shipping agents in production

02

Tech coaching

For teams shipping to production

Working with your team on how they build — TDD, BDD, clean code, continuous delivery — and fixing the problem rather than the symptom. These practices matter more now as AI is writing most of the code, not less.

  • Working alongside your team, not presenting at them
  • The practices that keep quality up as the pace goes up
  • Habits that stay after I leave

03

1:1 mentoring

For engineering leaders and senior engineers

Regular sessions to unblock your teams and think through the calls you’re not sure about. Bring whatever is actually bothering you — an agent that keeps going sideways, a roadmap you want a second opinion on, a team that has quietly stopped reviewing what the model writes.

  • Roadmap and agent-programme reviews between sessions
  • Practical guidance on tools, patterns, and guardrails
  • Direct feedback on what your teams are building

04

AI workshops

For product and engineering teams

A few days with your engineers, working in your own codebase instead of a demo project. They leave knowing how to test and review AI-written code — the way I wrote it up in Test-First Copilot, taught to real teams since.

  • Live coding in your real codebase, on your stack
  • Test-first patterns that hold when an agent writes the code
  • Quality guardrails for review, CI, and merging agent output

Who this is for

Does this sound like you?

If you recognise yourself in more than one of these, we should talk.

You are trying to work out where AI actually belongs in your business, and what is worth building versus buying

Your team ships to production every week, and you want AI in that process without quality quietly slipping

You are the one making the AI calls for your teams, and you would rather talk them through with someone who has made them before

You are rolling out Copilot, Claude Code, or Cursor and want it to stick beyond the first month

You have a large legacy codebase and a nervous feeling about pointing AI at it

How we would work

Start small, then make it stick.

A few people using AI well is not the same as your company getting better at it. Here is how we close that gap.

01

Discovery

A free fitment call where we map your AI goals, your current stack, how your teams work today, and what you are actually worried about. If I am not the right person, I will tell you.

02

Pilot

We pick one team and one real piece of work — small enough to move quickly, real enough that the result means something. We agree upfront what better looks like and how we will know.

03

Scale

What worked in the pilot becomes training, working agreements, patterns, and guardrails for the wider organisation, including the engineering practices that make the AI part safe.

04

Measure & adjust

We look at what actually changed in your delivery and your code quality, correct what drifted, and hand the practice to your own people so it does not depend on me being in the room.

Track record

Why me.

The AI part is recent. The engineering part is not.

  • Author of Test-First Copilot and two further books on test-driven development
  • Two published papers — Agentic Code Surgery for Brownfield Systems, peer-reviewed on Zenodo (2026), and Real Productivity, Rare Pushbacks, on putting AI agents into real product teams
  • Contributor to Google’s Gemini CLI and Microsoft’s VS Code Copilot Chat, author of spec-tracer, and many more
  • Cofounder and Director at Ampyard, where I do the AI consulting and tech coaching
  • Built and shipped two AI SaaS products end to end

Let’s talk about your team.

Book a free fitment call. We’ll go through where you are with AI, what your engineering practices look like underneath it, and whether I’m the right person to help you.

Book a free fitment call