AI Research Agent

A simple LangChain-based agent that conducts research, generates content, and saves findings to structured documents.

Overview

This project implements a simple, focused research assistant using LangChain and large language models. The agent can search the web, query Wikipedia, generate comprehensive documents, and automatically save research findings to files.

Features

• Multi-source Research: Combines web search and Wikipedia data
• Structured Output: Returns findings in a standardized format
• File Storage: Automatically saves research documents with formatting
• Error Handling: Robust error management for network and API issues
• Agent-based Architecture: Uses LangChain's agent framework for complex tasks

Requirements

• Python 3.8+
• Dependencies listed in requirements.txt

Installation

1. Clone the repository:

   git clone [repository-url]
   cd AI-Agent
   

2. Create a virtual environment:

   python -m venv venv
   

3. Activate the virtual environment:

   • Windows: .\venv\Scripts\activate
   • Unix/MacOS: source venv/bin/activate

4. Install dependencies:

   pip install -r requirements.txt
   

5. Set up environment variables: Create a .env file with your API keys:

   OPENAI_API_KEY=your_openai_key
   GOOGLE_API_KEY=your_google_key
   # Add other API keys as needed
   

Usage

Run the main script:

python main.py

Enter your research query when prompted. The agent will:

1. Gather information from the web and Wikipedia
2. Generate a structured research document
3. Save the results to research_output.txt

Project Structure

• main.py: Core application with LangChain agent setup
• tools.py: Custom tools for search, Wikipedia, and file operations
• requirements.txt: Project dependencies
• research_output.txt: Generated research documents

Technical Details

Components

• LangChain: Framework for LLM applications and agents
• LLM Integration: Supports OpenAI, Google Gemini, and Anthropic models
• Pydantic Models: Type validation for structured output
• DuckDuckGo Search: Web search integration
• Wikipedia API: Structured knowledge source

Output Format

Research documents are saved with the following structure:

• Research topic and summary
• Detailed content
• Research process documentation
• Sources and references
• Tools used

Customization

You can modify the agent's behavior by:

1. Changing the LLM provider in main.py:

   llm = ChatOpenAI(model="gpt-4")  # Change to your preferred model

2. Adjusting the system prompt for different instructions

3. Adding new tools to tools.py for additional capabilities

Limitations

• Depends on third-party API access and rate limits
• Research quality depends on the underlying LLM capabilities
• Web search results may vary in quality and relevance

License

MIT License