AI Agents with LangGraph

Welcome to the AI Agents with LangGraph project, developed as part of the DeepLearning.AI course. This project explores how to build stateful, multi-step AI agents using LangGraph, an innovative framework for building agentic workflows on top of LangChain.

📖 Project Overview

LangGraph extends LangChain by enabling cyclic, stateful computation using a graph-based model. This project demonstrates how to:

• Create AI agents that reason through complex tasks.
• Manage memory and state transitions between agent steps.
• Use LangGraph to define conditional logic and loops.
• Build interactive and adaptive workflows with LLMs.

🚀 Features

• Agent creation with memory and tool use.
• Directed graph logic for stateful execution.
• Integration with LangChain tools and agents.
• Demonstration of multi-turn task completion.
• Fine-grained control over flow and branching.

🛠️ Tech Stack

• LangGraph – Agent orchestration.
• LangChain – LLM interface and tools.
• Python – Core logic and flow control.
• OpenAI / Hugging Face – LLM providers (optional).

📅 Use Cases Demonstrated

• Multi-step task execution (e.g., writing, planning, summarizing).
• Tool-augmented agents (e.g., using search, calculators).
• Stateful agent loops (e.g., retrying tasks, reflecting).

📄 Getting Started

1. Clone the repo:

git clone https://github.com/your-username/langgraph-ai-agents.git
cd langgraph-ai-agents

2. Install dependencies:

pip install -r requirements.txt

3. Run the example:

python agent_demo.py

🌐 Resources

• LangGraph Documentation
• LangChain Framework
• DeepLearning.AI Course

---

Feel free to fork this repo and experiment with your own agentic workflows!