LangGraph Agents Project

This project demonstrates the use of LangGraph and LangChain to build various AI agents for document drafting, mathematical operations, and PDF processing using OpenAI models.

Features

• Document Drafter Agent: Interactively helps users draft, update, and save documents using natural language.
• Math Agents: Perform arithmetic operations (addition, subtraction, multiplication, division) via conversational interface.
• PDF Agent: Loads and processes PDF files, enabling AI-powered analysis and interaction with document content.
• Streaming and Tool Integration: Uses LangGraph's streaming and tool node capabilities for dynamic, multi-step workflows.

Project Structure

• main.py: Entry point for running agents.
• 5_agent_1.py, 6_agent_2.py, 7_agent_3.py, 8_agent_4.py, 8_1_agent_4.py, 9_agent_5.py: Example agent scripts.
• 1_first-graph.ipynb, 2_second-graph.ipynb, ...: Jupyter notebooks demonstrating agent usage.
• logging.txt: Log file for agent runs.
• pyproject.toml, uv.lock: Dependency and environment management.

Requirements

• Python 3.8+
• OpenAI API key (set in .env)
• Required packages:
  • langchain-core
  • langchain-openai
  • langgraph
  • langchain-community
  • python-dotenv
  • pypdf
  • IPython
  • openai
  • uv (for dependency management)

Installation

1. Clone the repository.
2. Create a virtual environment:

   python -m venv .venv

3. Activate the environment:
   • Windows: ./.venv/Scripts/activate
   • macOS/Linux: source .venv/bin/activate
4. Install dependencies:

   pip install -r requirements.txt

   Or, if using uv:

   uv pip install -r requirements.txt

5. Set your OpenAI API key in a .env file:

   OPENAI_API_KEY=your-key-here
   

Usage

• Run a specific agent script:

   python 8_agent_4.py

• Interact with the PDF agent: Ensure Stock_Market_Performance_2024.pdf is present in the project directory, then run:

   python 9_agent_5.py

• Jupyter Notebooks: Open and run the .ipynb files for step-by-step demonstrations.

Example

python 5_agent_1.py
# Enter your message when prompted to interact with the agent.

Troubleshooting

• If you see an error like 'pypdf' package not found, install it:

   pip install pypdf 

• For missing modules, ensure all dependencies are installed and your virtual environment is activated.