Emotion-Aware Chatbot with LangGraph

This project is an emotionally intelligent chatbot built using Langgraph, and Google's Gemini API via langchain-google-genai.

It routes user input to different response styles:

• 🧠 Logical responses (e.g., for questions or factual queries)
• 💬 Emotional support (e.g., for personal struggles or feelings)

---

🚀 Features

• 📥 Message classification using structured LLM output
• 🔁 Dynamic routing via LangGraph
• 🤖 Multi-agent behavior: "therapist" and "logical assistant"
• 🌐 Uses Gemini 1.5 Pro / Flash via LangChain integration
• 🌱 Built with Python, LangGraph, and LangChain

---

🧪 Example Use Cases

> Message: I'm feeling anxious and overwhelmed today
Assistant: I'm really sorry you're feeling this way. Can you tell me more about what’s been bothering you lately?

**# Setup Instructions**

# Clone the repo
git clone https://github.com/YOUR_USERNAME/langgraph-project.git
cd LangGraph Project

# Create virtual environment
python -m venv .venv
source .venv/Scripts/activate   # On Windows

# Install dependencies
uv add python-dotenv langgraph langchain ipykernel
uv add langchain langchain-google-genai google-generativeai python-dotenv

# Setup Environment Variables
Create a .env file with your Gemini API key:
GOOGLE_API_KEY=your_google_api_key_here

# How to Run
python main.py

# Architecture Overview
       User Input
           ↓
    [Classifier Node]
           ↓
      [Router Node]
    ┌───────┴────────┐
    ▼                ▼
Therapist Agent   Logical Agent