Created as a submission for the 2023 VT E-Club Startup Sprint - Made from start -> finish in 4 days
Check out the quick tutorial on how it works: https://youtu.be/n1GBjq30vQo
UPDATE: This project was developed and released prior to Cursor and Github Copilot's codebase search feature realeases. Codebase embedding and retrieval has since become an officially integrated feature in all leading AI-first IDEs
This project introduces a novel translation mechanism specifically designed for embedding codebases within Python, enabling large language models such as OpenAI's GPT-4 to much more accurately semantically search and interpret source code from various programming languages.
Allows users to rapidly learn and build off of any public these repositories by adding them to the knowledge of the language model.