It seems that the next sprint will include an effort to connect our database with LLM.<p>while researching this, I noticed there was a lot of buzz last year. but now it is more quiet on that front.<p>Can you share how it worked for you? especially from the user perspective?<p>Thanks,
I recommend building an intermediary data access layer like a graphQL API that the LLM consumes, otherwise you run the risk of someone successfully tricking the LLM to drop all or part of the DB via a raw input-to-SQL step.