Dave Hudson · Contract AI Product Engineer
Core stack
React + TypeScript
Working location
Remote UK
Team fit
Small product teams

Products and AI workflows, built around human judgement.

I join small teams to shape ideas and ship production React and TypeScript. Agents do the repetitive work; scope, tests and approval keep people in control.

The delivery loop starts with human intent and context. An agent works within that scope and runs tests. A person reviews the evidence before the outcome ships, or sends the work back for another pass.

Senior delivery  ·  Remote UK contracts

20+ years
Shipping production software
Small product teams
Product shaping and engineering
Remote UK
Through Applification Ltd

SELECTED CLIENT OUTCOMES

Commercial work under real constraints.

Greenfield builds, architectural resets and production systems for startups, government and health technology teams.

  • LOGICALLY

    Days to minutes

    Intelligence v2 rebuilt for production

    As Principal Engineer, I led the team, architected and built Intelligence v2 with one other engineer. Routine UI changes fell from days to minutes, and the team released several times per day.

    REBUILD TO RELEASE
    6 months
    UI CHANGE TIME
    Days → min
    PRODUCTION RELEASES
    Several / day
  • ERUPTIV4 months

    Greenfield frontend live

    Designed and built the entire TypeScript and Next.js frontend for Client Server's recruitment platform against an existing API in three months. It reached production after a fourth month of release work.

  • PEPPY HEALTH£12m ARR

    Component architecture rebuilt

    Led a two-person senior frontend team, catalogued hundreds of components in Storybook and added a full Cypress end-to-end suite while the business scaled to £12 million ARR.

FLAGSHIP PRODUCT  ·  STORYLOOPS

In Development

A product map that coding agents cannot quietly ignore.

StoryLoops gives each task product context, then makes scope changes visible for approval before the map changes.

A StoryLoops product map pairs product work with an agent panel that waits for human approval before changing scope.

02  /  PLANTRY  ·  APPLE PLATFORMS

R&D

Meal planning that understands the household.

Plans the next 2 to 7 days around what the household likes, what needs using and what is in season. Shopping hands off to Reminders.

Plantry iPhone preview showing a three-day August meal plan and a prompt to plan tonight's dinner.
  • Plans around preferences, effort and what needs using
  • Learns from meals cooked, skipped or changed

DESIGNED, BUILT AND RELEASED

Open-source products.

Contexture gives Convex apps and coding agents one reviewed source model. Voiced pastes speech into the field you are already using.

CONTEXTUREOPEN SOURCE  ·  CONVEXLive

The visual source of truth for Convex apps.

Model tables, refs and indexes once, generate the files your app imports, and let coding agents propose constrained, reviewable changes through MCP.

AI-assisted engineering

AI is now part of how I engineer software.

I use agents to shape scope, gather context, implement, test and review. The work still ships on evidence and human approval.

See how I work with AI

CONTRACT ENGINEERING · PROJECT TEAMS

Need a senior product engineer, or a small team for the whole build?

I can join an existing team or assemble the product, design and engineering team for a complete build. Remote across the UK.

Discuss a contract