Holy Mole

See the cost of what's missing. And the cost of what is at risk.

Comment

Inspiration

Restaurant margins are thin. A single missing ingredient means I can't sell multiple high-value dishes. Traditional inventory tools just count items; they don't show the financial impact. I built a risk engine to visualize the cost of stockouts in real-time.

What it does

Holy Mole is an inventory dashboard that connects ingredients to revenue.

  • Traffic Light Monitoring: Flags items as Healthy, Warning, or Critical based on par levels.
  • Blast Radius Visualizer: Uses a graph to show exactly which menu items are blocked by a shortage.
  • Revenue Ticker: Calculates the exact revenue lost per hour due to missing stock.
  • Chaos Simulation: A button to simulate a rush and test the system's reaction.

How I built it

I used a hybrid tech stack.

  • Backend: Python and FastAPI handle the logic, using a recursive DFS algorithm to traverse the dependency tree.
  • Frontend: Next.js 14 and TypeScript with Tailwind CSS and shadcn/ui for the dashboard.
  • Visualization: React Flow renders the dynamic dependency graph.
  • Data: SQLite handles the transactional stock and usage data.

Challenges I ran into

  • Modeling the Menu: Mapping the many-to-many relationships between ingredients, sub-recipes, and dishes was complex.
  • State Management: Getting React Flow to update node colors instantly during the simulation required careful state handling.
  • CORS Errors: Connecting the Next.js frontend to the Python backend required precise configuration.

Accomplishments that I'm proud of

  • The Algorithm: Writing a clean graph traversal that identifies every affected menu item.
  • The UI: Creating a dashboard that looks like a mission control center rather than a spreadsheet.
  • The Resolution: The satisfaction of the restock button turning the entire graph green immediately.

What I learned

  • Graph Theory is Practical: Concepts like DAGs are perfect for modeling supply chain dependencies.
  • Revenue Context: Showing dollar amounts changes user behavior faster than showing quantities.
  • Full-Stack Sync: Keeping a Python backend and React frontend aligned requires rigorous API design.

What's next for Holy Mole

  • Predictive AI: Using sales history to predict rushes before they happen.
  • Supplier Integration: One-click ordering directly from the dashboard.
  • Multi-Location Support: Allowing franchise owners to monitor multiple kitchens.

Built With

Share this project:

Updates