Hey HN! Govscent is a project to do topic extraction and summarization for all bills going through the US congress, via ChatGPT, provided to the public for free. The name means "gov-scent", like, smelling the government. :)

We just passed 130k bills analyzed.

At the moment we only have bills going back to 1993, but soon we'll be adding all bills and amendments going back to the 1st congress.

Right now there is only text search of the topics, but soon you'll be able to jump to a specific bill, etc.

It's something I've been building off-and-on for the past few weeks. Only about an hour or so a week so it's still rough around the edges. My partner has helped too, as her first frontend-ish project. Contributions welcome! The parsing code to handle the responses from the LLM is very naughty, if you can stand it.

The data is also available for free, read-only via the API: https://govscent.org/api/

It all runs on a single Linode today so be kind. :)