Research Rabbit Hole

Research Rabbit Hole is an advanced AI agent system capable of conducting autonomous, deep research. It orchestrates multiple sub-agents to gather, analyze, and synthesize information into comprehensive reports.

🚀 How It Works

1. Scope & Clarify: The agent interacts with you to refine your research topic, asking clarifying questions to ensure the brief is precise.
2. Deep Research: Multiple sub-agents are deployed in parallel. They browse the web (via Tavily), read documents, and reflect on their findings to ensure deep coverage.
3. Synthesize: All gathered data is aggregated and processed into a structured, professional report with citations.

🛠️ Tech Stack

• LangGraph: Orchestrates the stateful, multi-agent workflow.
• FastAPI: Powers the backend service and API endpoints.
• Streamlit: Provides a modern, interactive chat interface.
• Tavily: Enables high-quality, AI-optimized web search.

📦 Quick Start

Prerequisites

• Python 3.11+
• uv (for dependency management)
• API Keys (OpenAI, Tavily, etc.) in .env

Installation

1. Clone the repository

   git clone https://github.com/HimathX/research-rabbit-hole.git
   cd research-rabbit-hole

2. Install dependencies

   uv sync

3. Run the Backend Service

   uv run src/run_service.py

4. Run the Streamlit App (In a new terminal)

   uv run streamlit run src/streamlit_app.py

🙏 Acknowledgements

This project is a fork of agent-service-toolkit by JoshuaC215. It has been customized for deep research tasks.