About this project
Description from the program archive
RAG-Based Chatbot for Enhanced EvalAI Challenge Support
Problem EvalAI struggles with providing consistent, timely, and accurate support for AI challenge participants and hosts, especially during peak periods and with frequently updated guidelines.
Solution I propose developing a Retrieval-Augmented Generation (RAG) chatbot that integrates seamlessly with EvalAI to offer instant, context-aware support.
Key Components 1. Document Ingestion Pipeline: Process and version diverse documentation formats 2. Vector Database (Milvus): Efficient storage and retrieval of embeddings 3. RAG Engine: Hybrid retrieval and context-aware answer generation 4. Web Interface: Intuitive chat UI integrated with EvalAI's platform
Deliverables 1. Production-ready RAG chatbot integrated with EvalAI 2. Comprehensive documentation and test suite 3. Contribution guidelines for future enhancements
This project aims to significantly improve EvalAI's support system, ensuring quick, consistent, and up-to-date responses for all users.