Skip to content

This repository contains various implementations of AI agents using LangGraph, demonstrating different patterns and use cases for building intelligent, tool-enabled conversational systems. Each project explores different aspects of agentic AI development, from basic chatbots to sophisticated RAG (Retrieval-Augmented Generation) systems.

Notifications You must be signed in to change notification settings

horus-bot/AgenticAI

Repository files navigation

LangGraph Agentic Models Practice

Author: Harsh Srivastava

Handle: horus-bot

Purpose: Practice repository for LangGraph agentic model implementations

📖 Overview This repository contains various implementations of AI agents using LangGraph, demonstrating different patterns and use cases for building intelligent, tool-enabled conversational systems. Each project explores different aspects of agentic AI development, from basic chatbots to sophisticated RAG (Retrieval-Augmented Generation) systems.

🏗️ Project Structure 🚀 Key Features

  1. Mathematical Tools Agent (toolsChatbot.ipynb) Performs arithmetic operations (addition, subtraction, multiplication) Demonstrates tool calling with LangGraph state management Uses ChatGroq with Llama-3.3-70b-versatile model

  2. RAG System (rag/ragModel.ipynb) Document-based question answering PDF processing with vector embeddings ChromaDB for persistent vector storage Contextual search with citation support

  3. Email Automation Agent (emailAgent.ipynb) Automated email composition and sending SMTP integration with Gmail Natural language email generation

4 Conditional Routing (2nd_level_contion.ipynb) Multi-level decision trees Complex workflow orchestration Dynamic routing based on user input

🛠️ Technology Stack Component Technology Framework LangGraph LLM Provider ChatGroq (Llama-3.3-70b-versatile) Vector Database ChromaDB Embeddings HuggingFace (all-MiniLM-L6-v2) Document Processing PyPDF, LangChain Email SMTP, aiosmtplib

📋 Installation Prerequisites Environment Setup Create a .env file in the root directory:

🎯 Quick Start

  1. Mathematical Tools Agent
  2. RAG System
  3. Email Agent 🔧 Core Patterns State Management All agents use consistent state management:

Tool Integration Tools are defined using LangChain's @tool decorator:

Conditional Routing Agents use conditional edges for decision making:

📊 Example Workflows Basic Chatbot Flow Tool-Enhanced Agent Flow RAG System Flow 🎓 Learning Objectives This repository demonstrates:

LangGraph Fundamentals: State graphs, nodes, and edges Tool Integration: Seamless LLM-tool interaction State Management: Message handling and conversation flow Conditional Logic: Dynamic routing and decision making Error Handling: Robust system design RAG Implementation: Document-based AI systems 🔍 Key Concepts Explored State Graphs: Building complex AI workflows Tool Calling: Extending LLM capabilities Conditional Routing: Dynamic decision making Vector Databases: Efficient document retrieval Multi-turn Conversations: Context preservation Error Handling: Graceful failure management

🚨 Common Issues & Solutions Module Import Errors

API Key Issues

Ensure .env file is in the correct location

Verify API key validity

Check environment variable loading

Tool Execution Errors

Verify tool call arguments

🤝 Contributing This is a personal learning repository. Feel free to:

Fork and experiment

Submit improvements

Share feedback and suggestions

Add new agent patterns

📚 Resources LangGraph Documentation LangChain Documentation ChatGroq API ChromaDB Documentation

📄 License Educational and research use. Please follow respective API terms of service.

Happy Coding! 🚀 Building the future of AI agents with LangGraph

About

This repository contains various implementations of AI agents using LangGraph, demonstrating different patterns and use cases for building intelligent, tool-enabled conversational systems. Each project explores different aspects of agentic AI development, from basic chatbots to sophisticated RAG (Retrieval-Augmented Generation) systems.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published