Neural Food Waste Predictor

Estimate food waste in real time from a browser-only dashboard.

This hosted demo keeps the product accessible from any computer by running a deterministic, in-browser version of the repo's waste logic and analytics flow.

Historical view

Recent waste pattern snapshot

The public demo synthesizes a stable historical window using the same operational factors as the training script.

Model framing

What drives waste in this demo

  • Meals served: higher meal volume increases the waste baseline.
  • Day of week: weekend service adds extra buffet-style waste pressure.
  • Temperature: hotter days increase spoilage risk.
  • Special event: events create the biggest over-preparation jump.

Live prediction

Adjust the operating conditions

Predicted output

Estimated waste today

60.0 kg