Talem AI is a RAG application. This codebase holds the code for the REST API mircoservice hosting the backend functions. It's main purpose is to act like a Q&A chatbot who provides answers to prompts about different college programs. To gain specialized context within this domain, we vectorize college information brochours and through Langchain, this information is retrieved based on it's similarity to the user's prompt and given as context to the prompt into Groq REST API.

Python is the programming language of choice for this project

Langchain abstracts a lot of the RAG logic, allowing for faster development times.

AstraDB allows for the storage of the vector embeddings (context) which are retrieved by the different materials we want our LLM to be well-versed in.

All AI needs like LLM and vector embeddings is taken care by this module.

FastAPI allows us to build the REST API which will allow the frontend interface to request information through their query. Development times are fast and it was easy to set up. Response times are also fast.