About — Agentic People

I'm a running coach, org designer, systems thinker, and AI builder. Agentic People is where I explore a practical question: as AI becomes more capable, how do people become more capable too?

Running Coach

I've been coaching runners for over 16 years, with more than 30 athletes trained across that time. At the peak of Mile Pursuit Coaching, I was working with 77 athletes concurrently — a number that required building real systems for memory, communication, and pattern recognition.

My coaching philosophy is patient, long-game, and fundamentally permission-giving. I'm not interested in the coach as authority figure. I'm interested in creating conditions where athletes can hear themselves better — where they develop their own sense of what's working, what isn't, and why. The goal is athletes who understand their own training, not athletes who are dependent on me to tell them how to feel.

I also have 25 years of my own training data, which gives me a specific kind of patience with the long arc of development. A lot of what looks like underperformance in year two looks like foundation in year six.

Org Design

I spent years working on organisational design and complexity at The Adaptavist Group (TAG), a global software and services company. That work was grounded in systems thinking — the idea that organisations produce the behaviour they produce because of how they're structured, not because of who's in them.

The tools I kept reaching for were facilitation, structural framing, and what I'd broadly call sensemaking: helping groups understand what kind of problem they're actually in, before they start solving it. Cynefin, sociometry, viable systems model — frameworks that help distinguish between complicated problems (where expertise helps) and complex ones (where you need to probe first).

AI Builder

I'm a hands-on AI builder. I use Claude Code heavily for development work — pipelines, home automation systems, data infrastructure. I've built a significant body of personal tooling: a home bridge that connects Hue, WiiM, Ring, and a dozen other systems; training data pipelines for coaching; prompt engineering infrastructure that's evolved into something like an operating system for AI-assisted work.

I'm not just a user of AI tools — I think a lot about how they're architected, what makes them reliable, and what happens at the boundary between AI judgment and human judgment.

Agentic People

Agentic People is the name I've given to the thing I'm most interested in right now: people using AI without giving up judgment, authorship, or responsibility.

Agency is not the same as autonomy. It is the capacity to notice what is happening, make meaningful choices, and act with consequence. AI can strengthen that capacity by holding memory, surfacing patterns, and widening the range of possible moves. It can also quietly weaken it when people outsource interpretation and accept generated answers as decisions.

I come at this from three directions: coaching, where agency means helping someone hear and direct themselves; org design, where structures determine who can act and with what information; and building, where the boundary between human judgment and machine execution becomes concrete. The aim is not humans versus agents. It is people becoming more agentic together.