Published Paper Details:

The problem of turning what people say into something computers can understand, like SQL is still an issue for people who are not tech savvy and want to work with complicated databases. Usually, people have used systems that're very strict or models that are good at learning, like LSTMs but these often take a long time to work and are not very flexible. This paper is about BaseMind, a way of talking to databases using natural language that makes it easier to get data by using the LangChain system and the GROQ engine. By putting a user interface on the front and a strong database on the back the system can turn what people say into SQL quickly. Tests show that using the hardware from GROQ and making the questions clearer makes the system work much faster and still gives accurate results even with complicated questions that involve multiple tables and sums. BaseMind is a way to make data available to everyone. It helps to close the gap, between what people say and what computers can understand.