The website

I maintain FOAG's website: the ordinary, unglamorous kind of volunteering that keeps a small charity's public face working. It's not the interesting part of the story, but it's the part that has to happen reliably.

The prototype

The more interesting part is a mobile prototype I'm developing to help farmers identify crop problems and get relevant guidance: point a phone camera at a struggling plant and get something useful back, rather than nothing at all. Text-to-speech is a deliberate part of the design, not an afterthought: a lot of the people this needs to work for can't read, so a tool that only outputs text isn't actually a tool for them.

I intend to travel to Uganda in 2027 to test the prototype first-hand: to listen to farmers, watch the real constraints they work under, and change the software around what people actually need rather than what seemed sensible from a desk in the UK. Software built for a context you haven't stood in tends to be wrong in ways that are only obvious once you're there.