Jan 2026 - Feb 2026
AI Agent
Built an intelligent document assistant that allows users to chat with a curated library of PDF books to get instant, evidence-based answers. Implemented a Retrieval-Augmented Generation (RAG) system to ensure AI responses are grounded in real data. Features include a secure Admin dashboard for managing documents, role-based authentication (RBAC), and a modern UI with fluid animations.
AI Agent is an intelligent document assistant built for querying a curated library of PDF books through natural language chat. Instead of giving generic answers, the system retrieves relevant document chunks and uses them as context so responses stay grounded in the uploaded material.
The project includes a secure admin workflow for managing documents, role-based access control, and a modern user interface for smooth document conversations. The backend coordinates ingestion, embeddings, vector retrieval, and Gemini-powered response generation.
Key Features
- Chat with PDF documents using Retrieval-Augmented Generation
- Admin dashboard for document management
- Role-based authentication and access control
- Vector search over embedded document chunks
- Gemini integration for grounded AI responses
- Clean conversational UI with fluid interactions
Tech Stack
Next.js, TypeScript, Node.js, Python FastAPI, Google Gemini, Mongo Vector DB, RAG