Multi-Source Research Agent

This project implements a smart research agent using LangGraph powered by Google Gemini. The agent can perform parallel searches across multiple sources (Google, Bing, Reddit), analyze the results, and synthesize a comprehensive answer.

Features

• Multi-Source Research: Queries Google, Bing, and Reddit simultaneously for diverse perspectives.
• Intelligent Analysis: Uses LLMs to analyze search results from each source independently.
• Reddit Deep Dive: Specifically fetches and analyzes reddit posts and comments for community sentiment and detailed discussions.
• Synthesis: Combines insights from all sources into a final, well-structured answer.
• File Output: Saves the final synthesized response to a Markdown file (final_response.md).

Prerequisites

• Python 3.10+
• API Keys for:
  • Google Gemini (Google GenAI)
  • Bright Data Api

Installation

1. Clone the repository:

   git clone https://github.com/Luckman-Khan/Multi_Source_Research_Agent.git
   cd Multi-Source_Research_Agent

2. Install dependencies: It is recommended to use a virtual environment.

   # using uv or pip
   uv sync 
   # OR
   pip install -r requirements.txt

   (Note: Ensure you have langgraph, langchain, langchain-google-genai, praw, python-dotenv and other necessary packages installed.)

3. Set up Environment Variables: Create a .env file in the root directory and add your API keys:

   GOOGLE_API_KEY=your_google_api_key
   BRIGHTDATA_API_KEY=your_bright_data_api_key

Usage

1. Run the Agent: You can run the agent via the provided Jupyter Notebook main.ipynb.

   Open the notebook and run all cells. The last cell initiates the run_chatbot() function.

   run_chatbot()

2. Interact:

   • The agent will prompt you to "Ask me anything".
   • Type your research question (e.g., "Current state of AI agents").
   • Type exit to quit.

3. Output:

   • The agent prints the progress of searches and analyses to the console.
   • The final synthesized answer is printed to the console.
   • The final answer is also saved to final_response.md in the project directory.

Project Structure

• main.ipynb: The main entry point containing the LangGraph definition and the chatbot loop.
• web_operations.py: Contains functions for performing Google/Bing searches and Reddit API interactions.
• prompts.py: Stores the prompt templates used for analysis and synthesis.
• snapshot_operations.py: Helper functions for managing efficient state or browsing snapshots, if applicable.
• final_response.md: The output file where the latest answer is saved.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.