🚀 PolyRAG: Multi-Agent Retrieval-Augmented Generation Framework

PolyRAG is a modular, multi-agent RAG (Retrieval-Augmented Generation) system designed to intelligently process and query diverse data sources such as documents, spreadsheets, and images.

It combines specialized agents with a central coordinator to deliver accurate, context-aware responses across multiple data modalities.

---

✨ Features

• 🧠 Multi-Agent Architecture

  • Document Agent (text + Word docs)
  • Excel Agent (structured data)
  • Image Agent (visual inputs)
  • Aggregator & Coordinator

• 🔍 Advanced Retrieval Pipeline

  • Embeddings + Vector Store
  • Context-aware query routing

• 🧪 Evaluation Framework

  • Automated benchmarking with test datasets
  • JSON + Markdown evaluation reports

• 💾 Memory System

  • Persistent conversational context
  • Improved response consistency

---

⚙️ Installation

git clone https://github.com/sidharth-vijayan/PolyRAG.git
cd PolyRAG
pip install -r requirements.txt

---

▶️ Usage

Run the main app:

python app.py