Fallow
A local events catalog that continuously checks its information against more than 1,000 sources, keeping listings useful without relying on manual upkeep.
More than 1,000 registered sources
The problem
Local events are scattered across venue websites, community calendars, government pages, and event platforms. Information changes constantly: recurring events move, festivals take a year off, and links stop working. A catalog that depends on someone checking every listing quickly becomes outdated.
I wanted to build one that could keep up.
The approach
- A broad source network. More than 1,000 registered sources contribute events to one searchable catalog
- Automated collection. Scheduled workflows gather and update listings, using different collection methods depending on the source
- Ongoing accuracy checks. The system checks live event information against its original sources and flags discrepancies
- A focused review process. Unusual cases are surfaced for review, so no one has to inspect every listing
- A private visitor experience. No accounts or personal information are required to browse and contribute
- Quality checks before release. Changes pass automated checks and a production build before they ship
The result
Fallow runs as an ongoing collection and maintenance system, with scheduled updates, source checks, and a review process for exceptions. As the source network grows, the catalog can continue operating without requiring a person to maintain every listing.
Screenshots
These screenshots show invented demo data from a local run: 22 made-up events in made-up towns.
Stack
Next.js on Cloudflare, Supabase, and a local Node executor for scheduled work.
Evidence
More than 1,000 registered sources, kept accurate by a daily automated check of live event information against its original sources.
Links
Source: private repository, available on request.