Dataline

DataLine is an open source AI data analysis and visualization tool that lets people chat with their data. A user connects a database or data file, asks a question in plain language, and DataLine writes and runs the SQL, returns tables and charts, and lets the results be copied into reports. Queries and charts can be edited, saved and rerun, and charts refresh when their underlying query changes. The project targets two groups: non technical people who want to explore data without writing SQL, and developers and analysts who want a fast text to SQL assistant for drafting queries and exploring unfamiliar databases. It connects to Postgres, MySQL, Snowflake, Microsoft SQL Server, Azure SQL Server and SQLite, and also queries CSV, Excel and sas7bdat files. It is privacy focused: everything is stored on the user's device, and by default the data itself is hidden from the language models it calls, a safeguard that can be switched off for data that is not sensitive. It runs as a downloadable executable for Windows, through Homebrew on Mac and Linux, or as a Docker image that adds optional authentication, and it serves its interface in the browser on a local port. DataLine is free and open source. The website publishes no paid plans; the only running cost is the language model provider the user configures. The repository README currently states that the project is looking for maintainers, and dashboards, triggers and a trainable knowledge base remain listed as roadmap items rather than shipped features, which is worth weighing before adopting it for production reporting.