🤖 Agentic AI Chatbot

FastAPI + Streamlit + Groq and OpenAI powered Agentic Chatbot

An end-to-end Agentic AI Chatbot that combines a FastAPI backend with a Streamlit frontend, supporting multiple LLM providers (OpenAI & Groq) with optional web search via Tavily. The project is designed with a clean, modular, production-ready architecture and can be run locally with a single command.

▶️ Agentic AI Chatbot – Live Demo

https://github.com/SHITICHY21/agentic-ai-chatbot/blob/main/Demo%20Chatbot.mp4

✨ Features

• 🧠 Agentic reasoning using LangChain & LangGraph
• 🔌 Multiple LLM providers:
  • OpenAI (GPT models)
  • Groq (LLaMA-3 models)
• 🌐 Optional Web Search (Tavily API)
• ⚡ FastAPI backend (REST API)
• 🎨 Streamlit interactive chat UI
• 🚀 One-command startup (Backend + Frontend together)
• 🔐 Secure API key handling using environment variables

🧱 Project Structure

agentic_chatbot/
│
├── agents/
│   ├── ai_agent.py          # LangGraph agent logic
│   ├── llm_provider.py      # OpenAI / Groq provider selection
│   └── tools.py             # Optional tools (web search, etc.)
│
├── app/
│   ├── route.py             # FastAPI routes
│   ├── model.py             # Pydantic request/response models
│   └── config.py            # Environment configuration
│
├── frontend/
│   └── streamlit_app.py     # Streamlit UI
│
├── main.py                  # FastAPI entry point
├── run_app.py               # One-command launcher (backend + frontend)
├── requirements.txt
├── .env.example
└── README.md

🛠️ Tech Stack

• Backend: FastAPI, Uvicorn
• Frontend: Streamlit
• LLM Framework: LangChain, LangGraph
• Models: OpenAI, Groq (LLaMA-3)
• Web Search: Tavily API
• Language: Python 3.11

📦 Installation

1️⃣ Clone the repository

git clone https://github.com/your-username/agentic-ai-chatbot.git
cd agentic-ai-chatbot

2️⃣ Create & activate virtual environment

python -m venv venv
venv\Scripts\activate

3️⃣ Install dependencies

pip install -r requirements.txt

▶️ Run the Application

🚀 One-Command Run (Recommended)

python run_app.py

Will start:

• FastAPI backend → http://127.0.0.1:8000
• Streamlit frontend → http://localhost:8501

👥 Contributors

Shiti Chowdhury

• 🎓 Department: Computer Science & Engineering (CSE), CUET
• 🔗 GitHub: SHITICHY21

Adnan Faisal

• 🎓 Department: Computer Science & Engineering (CSE), CUET
• 🔗 GitHub: AJFaisal002