Retail GenAI LLM

A GenAI LLM project for querying a retail t-shirt inventory database using LangChain, OpenAI, and MySQL.

Features

• Natural language to SQL query flow
• Few-shot prompting support
• Streamlit frontend for asking questions

Project Structure

• frontend/ - Streamlit app and helper modules
• langchain-db/ - Notebook experiments and SQL chain setup
• gpt_client.py - Shared OpenAI/LangChain client setup
• database/ - SQL schema script

Setup

1. Create and activate a virtual environment.
2. Install dependencies:
   • pip install -r requirements.txt
3. Add your API key in .env:
   • OPENAI_API_KEY=...

Database Setup

This project includes a dummy MySQL dataset for local testing:

• Script path: database/db_creation_atliq_t_shirts.sql
• Database name used by the app: atliq_tshirts

Load it into MySQL:

• mysql -u root -p < database/db_creation_atliq_t_shirts.sql

If needed, update DB credentials in frontend/main.py to match your local MySQL setup.

Run

• Start frontend app:
  • streamlit run frontend/main.py

System Design

Flowchart

Photos

Sample Questions

Try these in the Streamlit app:

• How many t-shirts do we have left for Nike in XS size and white color?
• How many white color Levi's shirts do we have available?
• How much is the total inventory value for all S-size t-shirts?