Skip to content

Latest commit

 

History

History
28 lines (23 loc) · 1.32 KB

README.md

File metadata and controls

28 lines (23 loc) · 1.32 KB
Raw

University of Cincinnati Chatbot ( Unibot)

Description

This chatbot assists students at the University of Cincinnati by providing accurate, real-time answers to their queries. Built with the Retrieval-Augmented Generation (RAG) framework, the chatbot pulls data from university knowledge bases and uses advanced natural language processing (NLP) through LLAMA-3 for intelligent responses. The system is designed for scalability and ease of use, combining a serverless architecture with Cloudflare Workers for the backend.

Features

  • Real-time, accurate responses using RAG and LLAMA-3 models.
  • Serverless backend using Cloudflare Workers.
  • Scalable and fast queries with Pinecone as the vector database.
  • PostgreSQL for storing persistent data such as FAQs and logs.
  • React-based frontend for a responsive user interface.

Technologies

  • Backend: Serverless architecture powered by Cloudflare Workers.
  • Frontend: React.js.
  • Database: PostgreSQL (for data storage) and Pinecone (as a vector database for semantic search).
  • Machine Learning: LLAMA-3 model integrated with the RAG framework for NLP.

Installation

Prerequisites

  • Node.js v14+
  • PostgreSQL
  • Cloudflare Workers CLI (wrangler)
  • Pinecone API access

Clone the Repository

git clone https://github.com/username/university-chatbot.git