grosbeak

Nee Jemison.

A single-file search engine, for small (100K-1M page) sites.

why?

Two reasons.

• I would like to build a repository of sites and papers that is searchable. I want to own the content and the index, and I want it to be easy to deploy/self-host.
• I want a search backend for LLM-based applications where I control the content.

requirements

• An S3-compatible object store
• SQlite or Postgres

The use-case is (essentially) single-user.