Faisty – Fastmail Mailbox as SQL Through UI and MCP
Faisty is a tool that mirrors Fastmail email accounts into a user-owned PostgreSQL database, enabling real-time access to messages through SQL queries or natural language searches. It uses embeddings for semantic search and supports integration with AI agents via SQL or MCP protocols. The service operates in real-time, indexing emails as they arrive without requiring model training or batch processing.
- ▪Faisty syncs Fastmail accounts with a PostgreSQL database in real-time using a live JMAP stream.
- ▪Messages are embedded for semantic search, allowing users to query emails in natural language or via SQL.
- ▪AI agents can access the indexed data through SQL or MCP without handling user credentials, using a read-only OAuth token.
Opening excerpt (first ~120 words) tap to expand
FAIST · faist.dev SECTION 01 · OPEN 0% F faist how it works interfaces pricing privacy docs sign in start free → f v0.4 · public beta · self-host ready your fastmail, indexed. every message in your fastmail account, mirrored into a SQL database you own, with embeddings for semantic search. natural language for humans. SQL for agents. start free → see how it works no credit card · MIT core · OAuth in 3 clicks data in your own postgres no model training, ever pg_dump & go FAIST / 2026 / FASTMAIL → SQL · EMBEDDED · REAL-TIME scroll v0.4 · beta · build 1184 02 / 06 · the index indexed · embedded · queryable your inbox, as a database that thinks. every message becomes a row. every row carries an embedding vector for similarity search. all of it real-time, all of it in postgres.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at faist.
Discussion
0 commentsMore from faist
