Project Description
Our project focus is to ideate on how we can leverage LLM and Generative AI to explore and discover data and their relationships from the vast data repositories like Atlas Catalog and Data Victoria.
Currently you need expertise to interact with these catalogs and domain understanding to be able to discover and analyze data in these repositories. This is a barrier for general users who are missing out on finding good insights from this public available shared data pool.
Generative AI like ChatGPT has made AI accessible to general users. Our objective is to apply this same experience to make the Atlas and Vic Gov data discovery and analysis accessible to general users.
Data Story
Discovering data from an data catalogue like Atlas is very hard. In a day and age where ChatGP has made AI accessible to everyone, we want to try make these catalogues discoverable and easily analysable by anyone.
is our architecture.
We are leveraging Large Language Models via LangChain on the catalogue and its data.
The data for our prototype is from the following locations.
During the discovery process, we analysed various steps as shows in the images below;
