LLMs are still not very good to understand csv files, esp if some data types are messed up.
one way to solve it is via a natural language query transformed into a sql query.
this project is a simple example how to query a postgres database using natural language. the sample database is hosted on supabase and contains swimming results for a local swim club from zionsville, indiana. i created a postgres database on supabase and merged several csv files there. this is my very first project in any kind of programming. don't judge too strictly please :-)