jagorczyk@pl:~$
$ cd ~/projects

igor@warsaw:~/projects/rag$cat README.md

./rag

RAG Document Management System

Multimodal document management with an AI assistant over text, PDFs, and images.

~/projects/rag/README

# about

A document platform where teams upload files and query them through a chat interface powered by retrieval-augmented generation.

Similarity search runs on PGVector; a knowledge graph stores entities, relations, facts, aliases, and identity-resolution data so people and concepts stay linked across documents.

Image workflows include OCR, visual analysis, face embeddings, and face recognition for recurring people.

# highlights

  • Chat-first interface over folders, conversations, and file previews
  • RAG + vector search + knowledge graph in one system
  • Vision pipeline for PDFs and images, not text-only dumps

# stack

# backend & ai

  • Java 17
  • Spring Boot 4
  • LangChain4j
  • PostgreSQL / PGVector
  • Ollama
  • OpenAI / Groq
  • Llama
  • Embeddings

# frontend

  • Next.js
  • React
  • Tailwind CSS

# documents & vision

  • Apache Tika
  • PDFBox
  • OpenCV
  • OCR

# infra

  • Docker

igor@warsaw:~/projects$ls

More projects