// David Hawley · Windsor, London commutable
20 years B2C product. Now shipping agentic AI systems at Eleo Studio.
Open to senior product and founding roles. London, hybrid or remote. Available within a week.
An agentic cycling trip planner.
Planning a multi-day ride means juggling routes, gradients, daylight and where to sleep across a dozen tabs.
Ran the discovery, designed the agent loop, built it. It turns a rough ambition into a day-by-day tour: routes, bike-friendly stays, weather, cafe stops. LLM-as-judge eval suite behind it.
A white-labelled recruitment engine.
Job boards return hundreds of near-misses and no sense of which roles are actually worth a reply.
Built and shipped a recruitment engine, then white-labelled it as a product: it pulls roles, scores each against a profile with an LLM, and ranks a shortlist with the reasoning attached. I run my own job hunt through it.
A native training app, Mac to head unit.
Structured indoor training means a subscription, a closed ecosystem and someone else's idea of a workout.
Built a native macOS training app in SwiftUI: ERG-mode structured workouts driven over ANT+, live power, cadence and torque, then deep post-ride analysis: mean-maximal-power curves, time in zone and a FIT export. It also runs on the bike, a companion Hammerhead Karoo app in Kotlin with custom on-device data fields.
Structured workshops and user interviews to find the real problem before a line of code.
A working slice in days, in front of real users, to learn what the spec missed.
Architect it, build it and put it in production. I write the code, not just the brief.
Instrument what matters, watch how it gets used, then cut and sharpen.
Open to senior product and founding roles. London, hybrid or remote. Available within a week.