~/vimal-jain
GenAI PlatformFeb 2026

PDF Summarization Platform

Distributed document processing platform handling up to 1,000 PDFs using hierarchical map-reduce AI summarization.

High-throughput PDF processing engine designed for large document workloads.

Uses a Node.js/TypeScript Express backend paired with BullMQ, Redis, Socket.IO, MinIO object storage, and MongoDB.

AI processing is offloaded to a Python FastAPI service running local Ollama LLM models with hierarchical map-reduce chunking.

Frontend built with Next.js featuring real-time progress updates via WebSockets.

stack

Next.jsTypeScriptExpressBullMQRedisSocket.IOMinIOMongoDBFastAPIOllama