Prakhar-Bhartiya/Agentic-Deep-Researcher

Agentic Deep Researcher

Agentic Deep Researcher is a powerful multi-agent system designed to perform deep web searches and synthesize structured insights using LinkUp and CrewAI. It provide comprehensive research capabilities, including deep web search, analysis, and structured response generation.

Features

• Deep Web Search: Perform both standard and deep web searches.
• Agentic Design: Orchestrates multiple agents using CrewAI for efficient task delegation and collaboration.
• LLM Integration: Powered by Deepseek R1-7B(LLM)
• Interactive UI: Built with Streamlit for an intuitive and user-friendly interface.
• MCP Server: Can be used as an MCP server for seamless integration with other systems.

Technologies Used

• LinkUp - Web search tool for deep and standard searches.
• CrewAI - Framework for agent orchestration and task management.
• Deepseek R1:7B - Large Language Model for natural language processing.
• Streamlit - Framework for building interactive web applications.
• MCP (Model Context Protocol) - For server integration.

Setup Instructions

Prerequisites

• Python 3.11 or higher
• LinkUp API Key (required for web search functionality)

Installation

1. Clone the repository:

   git clone https://github.com/prakhar-bhartiya/agentic-deep-researcher.git
   cd agentic-deep-researcher
   

2. Install the required packages:

   uv sync
   

Configuration

1. Obtain your LinkUp API key from LinkUp.
2. Set up the MCP server configuration in mcp_config.json:

   {
     "mcpServers": {
       "agentic_research": {
         "command": "uv",
         "args": [
           "--directory",
           "./agentic-deep-researcher",
           "run",
           "server.py"
         ],
         "env": {
           "LINKUP_API_KEY": "linkup_api_key"
         }
       }
     }
   }
   

   Replace linkup_api_key with your actual LinkUp API key.

Usage

Running the Application

To run the application, use the following command:

streamlit run app.py

Using as MCP Server

To use Agentic Deep Researcher as an MCP server, ensure the MCP configuration is set up correctly in mcp_config.json, then start the server using:

uvicorn server:app --host 0.0.0.0 --port 8000 --reload

License

This project is licensed under the MIT License - see the file for details.