AI-powered document understanding and conversational retrieval system built using LangChain, ChromaDB, FastAPI, Streamlit, and Groq LLMs.

Document Intelligence RAG Agent allows users to:

• Upload PDF documents
• Perform semantic search on document content
• Ask natural language questions
• Retrieve context-aware AI-generated answers

The system uses a Retrieval-Augmented Generation (RAG) pipeline to combine vector search with Large Language Models for accurate responses.

✅ PDF Upload & Processing ✅ Semantic Document Search ✅ Conversational AI Q&A ✅ Vector Embeddings using Sentence Transformers ✅ ChromaDB Vector Store ✅ FastAPI Backend ✅ Streamlit Interactive UI ✅ Groq LLM Integration ✅ Modern Glassmorphism UI ✅ Dark-Themed AI Interface

PDF Upload
    ↓
Text Extraction
    ↓
Chunking
    ↓
Embeddings Generation
    ↓
ChromaDB Vector Storage
    ↓
Similarity Search
    ↓
Relevant Context Retrieval
    ↓
LLM Response Generation

rag-agent/
│
├── app/
│   ├── services/
│   ├── utils/
│   ├── uploads/
│
├── frontend/
│   └── frontend.py
│
├── .streamlit/
│   └── config.toml
│
├── requirements.txt
│
└── README.md

git clone <YOUR_GITHUB_REPO_URL>
cd rag-agent

python -m venv venv

Activate environment:

venv\Scripts\activate

source venv/bin/activate

pip install -r requirements.txt

Create a .env file:

GROQ_API_KEY=your_api_key_here

uvicorn app.main:app --reload

Backend runs on:

http://127.0.0.1:8000

streamlit run frontend/frontend.py

Frontend runs on:

http://localhost:8501

Add your screenshot here

README_images/app_preview.png

Example Markdown after adding screenshot:

![App Preview](README_images/app_preview.png)

1. Upload PDF document
2. System extracts and indexes content
3. Ask questions in chat
4. AI retrieves relevant chunks
5. LLM generates contextual answer

• Multi-document support
• Conversation memory
• Authentication system
• Cloud deployment
• Source citation highlighting
• OCR support for scanned PDFs
• Hybrid search (BM25 + Vector)

Summarize this document
What are the key findings?
Explain the methodology section
What technologies are discussed?

Built by Anurag Wanwe

This project is for educational and portfolio purposes.