📚 Library Desk Agent

An AI-powered Library Management Assistant built with FastAPI, LangChain, and Streamlit.
It helps manage books, customers, and orders through natural language queries.

🚀 Features

• 🤖 GPT-powered intelligent assistant (LangChain Agent)
• ⚡ FastAPI backend with tool orchestration
• 🖥️ Streamlit interactive frontend
• 🗄️ PostgreSQL database with SQLAlchemy ORM
• 🔄 Auto tool chaining (find_books → restock_book → create_order → ...)
• 💬 Persistent chat sessions with database logging

🧩 Project Structure

GenAI_Assessment/
├── app/
│   └── chat_ui.py             # Streamlit frontend
├── server/
│   ├── main.py                # FastAPI + Streamlit launcher
│   ├── db.py                  # Database connection
│   ├── models/                # SQLAlchemy models
│   ├── agent/
│   │   ├── chains/            # LangChain agent logic
│   │   └── tools/             # Custom AI tools
│   ├── routes/                # FastAPI endpoints
│   └── config.py              # Environment settings
├── db/
│   └── seed.py                # Seed script to populate sample data
└── README.md

⚙️ Installation & Setup

1. Clone the Repository

git clone <your-repo-url>
cd GenAI_Assessment

2. Create and Activate Virtual Environment

conda create -n desk-agent python=3.11 -y
conda activate desk-agent

3. Install Dependencies

pip install -r requirements.txt

4. Configure Environment

Create a .env file in the project root:

APP_NAME="" 
APP_VERSION=""
OPENAI_API_KEY=""
POSTGRES_DB_URL=""
POSTGRES_DB_NAME=""
POSTGRES_DB_USER=""
POSTGRES_DB_PASSWORD=""
API_URL=""

Environment Variables:

• APP_NAME: Application name displayed in the UI
• APP_VERSION: Current version of the application
• OPENAI_API_KEY: Your OpenAI API key for GPT integration
• POSTGRES_DB_URL: Full PostgreSQL connection string
• POSTGRES_DB_NAME: Database name (e.g., library_db)
• POSTGRES_DB_USER: Database username
• POSTGRES_DB_PASSWORD: Database password
• API_URL: Backend API endpoint for chat requests

🌱 Database Setup

Automatic Table Creation

When you run the application for the first time, all database tables are created automatically through SQLAlchemy ORM migrations.

Data Seeding

To populate initial books, customers, and sample data:

cd ~/path/to/your/project
python -m db.seed

Manual Schema Setup (Troubleshooting)

If you encounter any database issues, you can manually create the schema using the provided SQL file:

psql -U your_username -d library_db -f seed.sql

This will create all necessary tables and constraints for the application.

▶️ Run the Project

You can run FastAPI and Streamlit together from one terminal:

python -m server.main

Once launched:

• FastAPI runs on → http://127.0.0.1:5000
• Streamlit UI runs on → http://127.0.0.1:8501

🧠 Example Queries (Test Cases)

You can test directly in the Streamlit chat or via API calls.

🔍 Find Book

{ "query": "Find books by Robert C. Martin" }

📦 Restock Book

{ "query": "Restock Clean Code book by 45" }

🛒 Create Order

{ "query": "Create an order for customer 2 for 2 copies of Clean Code" }

💰 Update Price

{ "query": "Update the price of Clean Code to 50 dollars" }

📑 Order Status

{ "query": "What is the status of order 2?" }

📉 Inventory Summary

{ "query": "Show me all books that are running low on stock" }

🧪 API Reference

Endpoint	Method	Description
/	GET	Health Check
/chat	POST	Send a query to the AI agent
/sessions	GET	List all chat sessions
/messages/{session_id}	GET	Retrieve chat history for a specific session

👨‍💻 Developed By:

Mohannad Hendi
Senior AI Engineer • Automation Specialist • Data Scientist