Applying AI to Cut Food Waste

First-year Applied Data Science & AI student Carlijn van der Vleuten built TasteScape, a web application aimed at cutting food waste in households.

“Users can upload photos of leftover fruits and vegetables, and the AI then identifies the items in the image and suggests recipes based on them, drawing from a database of over 8,000 recipes.”

How it was built

The project started with curating and structuring recipe data scraped from the internet, using Python web frameworks. Carlijn then built a computer vision model, collecting and augmenting thousands of photos to train it to recognise 44 different fruits and vegetables. Training time and database integration were both sticking points; she worked through them and finished with a round of user testing to refine the experience.

Reflection

What she is proudest of is seeing the idea actually materialise — and what she learned getting there. Most of the obstacles came down to trial and error, refining both the database structure and the interface based on feedback.

Next she wants to go further into data engineering, driven by an interest in merging personal interests with solutions that mean something.