Fallow is a tool for farmers leaving animal agriculture. It takes a farmer's location, current operation, and biggest concern about transitioning, then generates a personalized land assessment: what their soil and region can support, which crops are realistic given their existing infrastructure, and what their actual path to income looks like.
The live site is embedded below. If you're unable to view or use it, visit fallow.farm directly.
Fallow came out of a broader question I keep returning to: AI safety and animal welfare conversations spend a lot of time on governance and research, but not much on whether the resulting tools are usable by the people who'd actually need them. Farm transition support exists, but it's high-touch and hard to scale. Fallow is an attempt to lower the barrier to that first conversation, not replace the organizations doing the real work.
I designed every screen in Figma, then built the full site myself using Claude as a development partner, exporting working React and Tailwind code straight from the design rather than handing it to a separate engineering process. The assessment runs on the Claude API, pulling in real geocoded land data and generating a response tailored to each farmer's specific situation.
This project was less about a single clean interface and more about owning a build end to end: research, design, prompt engineering, frontend code, and deployment. It's a working answer to the question I started with, that design and AI together can make a serious resource feel approachable rather than overwhelming.
