A complete, dual-sided intelligent restaurant management solution built for the hackathon. This project features two main components working in tandem:
- Part 1: Owner Intelligence Dashboard (Streamlit) - A backend analytics suite for pricing, combo generation, and ML demand forecasting.
- Part 2: AI Revenue Copilot (React) - A 100% offline, WebGPU-powered Voice/Text Point-of-Sale (POS) interface for the restaurant staff.
A powerful analytics suite built in Python that ingests POS data to optimize menu pricing, identify customer segments, and forecast future demand.
- Menu Engineering (BCG Matrix): Automatically categorizes menu items into Stars, Plowhorses, Puzzles, and Dogs based on popularity and contribution margin.
- Smart Combo Generator: Analyzes historical co-purchase frequencies to build intelligent, high-margin combo offers (e.g., pairing Butter Chicken with Naan).
- Price Simulator & ML Forecasting: Uses Scikit-Learn Linear Regression to predict how price changes will impact sales volume and overall profit.
- Customer RFM Segmentation: Groups customers based on Recency, Frequency, and Monetary value to identify champions and at-risk patrons.
- Dynamic Upsell Engine: Automatically generates and updates an
upsell_config.jsonfile based on real-time data filters.
Prerequisites: Python 3.9+ installed on your machine.
- Open a terminal and navigate to the Streamlit project folder.
- Install the required Python libraries:
pip install streamlit pandas plotly scikit-learn numpy
- streamlit run app.py
- Node.js installed on your machine.
- A modern browser with WebGPU support (Google Chrome or Microsoft Edge recommended).
- Important: Ensure "Use graphics acceleration when available" is turned ON in your browser settings to allow the AI to access your GPU.
- Clone or extract this repository and navigate to the project folder.
- Install the required dependencies:
npm install
- npm run dev
- Ctrl + Left Click on the Localhost link(for example: http://localhost:5173)