Mai Shan Yun - Inventory Intelligence Dashboard

Overview

Mai Shan Yun is a real-time inventory intelligence dashboard designed for restaurants or food businesses. It helps optimize:

• Ingredient usage forecasting
• Cost analysis and high-spend contributors
• Shipment delay tracking
• Ordering efficiency (waste vs. shortage detection)
• Menu suggestions based on ingredient trends

The dashboard converts raw purchase, sales, recipe, and shipment data into actionable insights and alerts.

---

Key Features

Overview & Alerts | Inventory usage, reorder warnings, ingredient-level metrics

Forecasting | Moving average prediction for ingredient demand

Cost Optimization | Detects top cost drivers and highlights spending patterns

Shipments & Logistics | Tracks delivery delays and visualizes distributions

Ordering Efficiency | Identifies recurring waste/shortages and suggests optimal ordering

Trend Analysis & Menu | Clusters trending ingredients using KMeans and generates menu ideas

---

Dataset Requirements

Place the following CSV files in a folder named data/:

purchases.csv
sales.csv
recipes.csv
shipments.csv

Required Columns

File | Required Columns | purchases.csv | ingredient_id, quantity, date, (optional) unit_cost sales.csv | menu_item_id, quantity_sold, date recipes.csv | menu_item_id, ingredient_id, quantity_per_item, ingredient_name shipments.csv | ingredient_id, purchase_date, delivery_date

If unit_cost is missing in purchases.csv, the app will use quantity as a temporary cost proxy.

---

How to Run the App

1. Install dependencies

pip install -r requirements.txt

2. Run the Streamlit app

streamlit run mai-shen-yun.py

3. Open the dashboard

Check the terminal output for the local URL (usually http://localhost:8501) and open it in a browser.

---

Example Insights

Reduce waste | Ingredient "Chicken Wings" shows consistent over-ordering; suggest ordering 8 fewer shipments |

Low stock warning | Only 2 days of inventory remaining based on average usage |

Menu strategy | "Ginger Grilled - Savory glaze" generated from trending ingredient clusters |

---

Tech Stack

Frontend dashboard | Streamlit

Data processing | Pandas / NumPy

Visualization | Plotly Express

ML clustering | Scikit-learn (KMeans)

---

Repository Structure

mai-shen-yun.py        # main Streamlit app
/data                  # CSV files go here
README.md              # documentation

Credits

Developed by: Noah Brown, Archelaus Paxon, Alphonse Thomas

Built With

• python