This project implements an intelligent, code-generating agent powered by contextual Retrieval-Augmented Generation (RAG) and the LlamaIndex framework. Designed to read and deeply understand both natural language documentation (PDF, DOCX) and source code files (Python, TypeScript, etc.) as well as searching on the internet for necessary information, the agent leverages contextual retrieval to produce accurate and context-aware code outputs. Generated code is automatically written to the appropriate files created and managed by the agent within user's requests.

✨ Key Features

• Contextual RAG with LlamaIndex Utilizes LlamaIndex to ingest, index, and retrieve relevant chunks from documentation and code files. The system enables advanced reasoning over multi-format data to ensure precise and informed code generation.

• Multi-User Support Built with scalability in mind, the platform supports multiple concurrent users. Optimized retrieval and generation pipelines ensure fast, low-latency query handling across sessions.

• Facial Recognition Authentication Secure user authentication is implemented via facial recognition, enabling users to log in and manage their accounts safely. Each user can update personal information and retain their session state across interactions.