← Back to home
Docusight AI
How Docusight AI lets students and researchers chat with PDFs using a RAG pipeline built with FastAPI, OpenAI, and Pinecone.
Problem
Students and researchers needed a faster way to work through dense PDFs — extracting answers, summaries, and study guides without reading every page manually.
Approach
As Senior Full Stack Engineer, I built the backend with FastAPI and PostgreSQL, implemented a Retrieval-Augmented Generation (RAG) pipeline using Pinecone as the vector database and OpenAI's LLMs for the chatbot, and built the Next.js/Tailwind CSS/TypeScript frontend for interactive document conversations.
Stack
- FastAPI
- PostgreSQL
- OpenAI
- Pinecone
- Next.js
- Tailwind CSS
- TypeScript
Result
The platform lets users chat directly with their PDFs, get instant summaries, and generate study guides, delivering a seamless AI-powered study experience.