Built

What my AI team has built

A dated log of finished work, newest first: things that are done, not planned. Everything here was directed, checked and signed off by me.

Only work that's true and not confidential appears here. Client work is added only with the client's permission.

Log

  1. This website

    Rebuilt waynegibbins.com from scratch: copy, design system, self-hosted fonts, the pages on this site, fact-check and QA, with an AI team of five agents directed by me.

    WebsiteCopyDesign

  2. –

    A prospect-research pipeline

    Built a prospect-research pipeline over Companies House data and company websites, with accuracy gates at every stage: 7,187 companies narrowed to 284 well-matched businesses.

    Screenshot of the prospect tool's map view in dark mode. On the left, a filter panel with tick boxes for tiers A, B and C and for each trade, and a choice of centre town and distance. On the right, a map of south-west England with companies grouped into labelled clusters around Trowbridge, inside dashed distance rings, and buttons above it to send the companies in view to a queue or export them.
    The map view: filter by tier, trade and distance from a chosen town, then send the companies in view to the Next up queue or export them. Map data © OpenStreetMap contributors.
    Screenshot of the prospect tool's patterns view in dark mode: a heat map with trades as rows and seven signal columns, from Tech behind 6+ to Established 25+ years. Each cell shows a percentage and a count, in lighter green as the share rises, and a Columns control above the table reads "Public, 7 shown".
    The patterns view, with the columns about directors hidden. Each cell shows how many companies in a trade share a signal, such as rising headcount or customers who are slow to pay, out of those where it could be worked out. Columns can be shown, hidden or reordered.

    ResearchDataTools

  3. –

    Home-Checker: property reports across England

    Built and launched Home-Checker, a paid report on any address in England covering flood risk, crime, schools, energy rating, air quality and the local property market. Each report calls several public data services as it's made (police.uk, the Environment Agency, the EPC register and planning.data.gov.uk), on top of national datasets loaded in bulk, such as Land Registry sales and schools.

    ProductDataPublic APIs

  4. –

    Ecstasis: focus and planning, designed with ADHD in mind

    Built Ecstasis, an iOS app and web dashboard for focus and planning, designed with ADHD in mind and around its founder's view of ADHD as a superpower: AI turns spoken notes into small-step tasks, a hyperfocus timer with soundscapes, and sleep data from Apple Health with the user's consent.

    ProductMobileAI

  5. –

    FloodWatch UK and the tools that run it

    Built FloodWatch UK, a flood-monitoring service using live Environment Agency, Natural Resources Wales and SEPA data across England, Wales and Scotland, with a web app and an iOS companion app. Around it: machine-learning models trained on river-level and rainfall history from 186 gauging stations and tested blind on a winter they'd never seen, a market-research back office for scheduling and logging customer interviews and surveys, and admin tools for users, organisations, site permissions, newsletters and system health.

    ProductDataMachine learning

Want something like this for your business? Book a 30-minute call or see what I do.