The Sovereign Engineer Workshop

AI literacy for experienced engineers: staying in command of code you didn’t write

(Available as a one-day masterclass or a two-day bootcamp; in-person and remote)

Get in touch

AI coding tools are fast, fluent and tireless. They also amplify whatever judgment you bring to them. The skill that matters now is not prompting. It is designing, checking and answering for your collaboration with AI.

This workshop is based on my book The Sovereign Engineer: AI Literacy for Software Professionals. You work with a real AI coding agent on a codebase you know, and climb the book’s six levels of AI literacy: from understanding what kind of collaborator an AI is, to running parallel agents safely while you stay answerable for every decision you delegate.

It is highly interactive, designed with Sharon Bowman’s Training from the Back of the Room approach: short concept segments, then hands-on practice with real AI agents, debriefs and your own written conclusions. The one-day masterclass covers the core practices with hands-on work throughout. The two-day bootcamp covers the full progression, with more time on your own code.

Best for groups of around 12 to 24, working in pairs. Bringing early-career engineers too? The Sovereign Apprentice workshop pairs with this one: seniors learn to design the environment, and juniors learn to earn their sign-off within it.

Reach out to find out more

Workshop Flow

  • Session 1: Know Your Collaborator

    Understand what kind of collaborator an AI is

    Know what you know that it cannot

    Say what you mean, and check that you were understood

    Spot output that is plausible but wrong

    Exercises include: working with a real AI coding agent on a codebase you know

  • Session 2: Tests as the Contract

    Hold the agent to a contract, not to its own confidence

    Write tests first, as the contract with your agent

    Use mutation testing to expose tests too weak to hold it to account

    Exercises include: writing the contract before the agent writes the code

  • Session 3: The Living Harness

    Make your team’s unwritten rules something an agent can follow

    Turn conventions into context, constraints and automated checks

    Slice work into one-decision pieces

    Write specifications an agent can build from

    Exercises include: building a living harness on your own code

  • Session 4: Agents at Scale, Answerable Throughout

    Hand more work to agents without handing over your judgment

    Run parallel agents and standing loops safely

    Put proper checks and budgets in place

    Stay answerable for the decisions you delegate

    Own the choice of model

For experienced engineers and tech leads who already use AI coding tools, and who want to stay in command of the code they didn’t write

You’ll leave this workshop with:

  • ✺

    A living harness of context, constraints and automated checks, built on your own code, not left in your notes.

  • ✺

    Tests that act as a real contract with your agent, and specifications it can build from.

  • ✺

    Safe habits for running parallel agents and standing loops, with proper checks and budgets.

  • ✺

    A clear view of which decisions you can delegate, and how to stay answerable for them.