Mai Shan Yun Inventory Intelligence Dashboard

Inspiration

Restaurant inventory management is complex. We built this dashboard to transform Mai Shan Yun's raw operational data into actionable insights that help minimize waste and prevent shortages.

What it does

Our Inventory Intelligence Dashboard provides:

Real-time metrics tracking menu items, ingredients, weekly shipments, and active alerts
Sales trend analysis visualizing performance across product categories (Ramen, Fried Rice, Wings & Cutlets, Rice Noodles) over 6 months
Ingredient usage tracking identifying top-consumed items to optimize ordering
Demand forecasting using linear regression to predict future sales patterns
Cost analysis breaking down weekly ingredient costs with visual percentage indicators
Shipment scheduling monitoring delivery frequency and timing
Smart recommendations suggesting bulk buying opportunities and waste reduction strategies
Ingredient complexity matrix showing which menu items share ingredients for better inventory planning

How we built it

Backend (FastAPI + Python):

Processes monthly Excel sales data (May-October)
Analyzes ingredient usage and shipment CSVs
Implements forecasting algorithms using numpy polynomial fitting
Calculates weekly usage metrics across different shipment frequencies
Provides 10+ REST API endpoints serving dashboard data

Frontend (Vanilla HTML/CSS/JS):

Clean, professional interface using Inter font and modern design patterns
Consistant with Mai-Shan-Yun's overall brand style
Interactive Chart.js visualizations for trends, forecasts, and usage patterns
Tabbed navigation between Sales Trends, Ingredient Matrix, and Demand Forecast
Responsive grid layout adapting to different screen sizes
Real-time data fetching with error handling

Challenges we ran into

Frontend-Backend Integration: One of our biggest roadblocks was connecting the frontend to the backend API. We had to debug CORS issues, implement proper error handling for failed requests, and ensure data types from Python (numpy/pandas) serialized correctly to JSON for JavaScript consumption
Converting biweekly and monthly shipment frequencies into standardized weekly metrics
Handling inconsistent CSV formats and missing data with robust error handling
Balancing detailed analytics with a clean, scannable interface
Implementing accurate forecasting with limited historical data

Accomplishments that we're proud of

Built a production-ready dashboard that processes real restaurant data
Created actionable insights like identifying high-frequency items for bulk contracts
Designed an intuitive UI that doesn't require training to understand

What we learned

Restaurant inventory has complex patterns (weekly/biweekly/monthly cycles)
Visuals are crucial
Data quality matters
Proper API design and error handling are essential for seamless frontend-backend communication