Automating Data News through Semantic-to-Querying

Languages & Libraries
Python, SQL, Pandas, Postgres, Git, Docker

In collaboration with The Metrics For One Matter, our team looks to cut down on latency between querying data and writing news articles, through the creation of an automated application to construct scheduled SQL queries against a database of real-world metrics. Connected to a Postgres database, the tool pulls DDLs automatically, filters for promising insight through a customized LLM, and executes queries to find statistically accurate data.

Recently, our team just wrapped up our final coding sprints, and have finalized our spring push, which can be viewed on Github. Shoutout to Aser & William for being great project partners, and Mike for his mentorship throughout the entire process. We are thinking of continuing the project through the summer to improve system accuracy, fully automate the system to generate written feedback, and eventually push results to a public audience.