SQL Chat

Hey folks of PH, We are from Bytebase, a Database CI/CD tool for DevOps teams. From our experience working alongside DBAs and Developer teams, we know that writing SQL is a headache for most people, let alone good SQL. With the help of Generative AI, here we present SQL Chat, the ChatGPT-powered SQL client. It currently supports MySQL, PostgreSQL, SQL Server and we are gradually adding more databases. ✌️ To try it out, simply open https://sqlchat.ai from your browser (no sign-up required) 👀 Chat with the default SQL Chat bot or the sample database. You can also create a connection to your own database, then ask questions, query, and change data using natural language. 🔒 SQL Chat itself doesn't retain any data. Your database connection is stored locally in your browser. And SQL Chat only sends the database schema to the OpenAI API. No table data will be transmitted. 😄 Let me know what you guys think! Thanks. Tianzhou Chen | Co-Founder / CEO

I tried sqlchat.ai recently, finding it quite useful for my ad-hoc query needs. I am one of those who love and hate sql simultaneously, appreciating its helpful query standard while toutured by its endless details and dialect differences. So after the overwhelming discussions about ChatGPT and OpenAI API, I was wondering if there are any tools relieving me from memorizing sql details. The moment I heard about the release of sqlchat.ai, I gave it a try. It is much faster than ChatGPT, with response waiting time less than 500ms or so. After entering my query requirment in the natural language word order, it can really understand the exact purpose, returning federated query sentences with where, join, group by, order by, etc. This is quite amazing because the only thing I need to do now is to focus on what I want to know from my db, which is the very important ASK step of data analysis. Being free from sql details really improves work efficiency. Quite a revolutionary UX improvement.

SQL Chat As far as my use in these days, it is different from the previous toy-level products, with a higher degree of maturity, and unlike the previous toy-level products, it has a certain degree of database design capability and is able to perform auxiliary management work. I have designed an RBAC model, which can understand the association between table structures, and when generating test data, the field meanings and types can match the table structures well.