Melbourne student startup uses AI tool to help households tackle food waste

A class at Haileybury Berwick in Melbourne has teamed up to create a new startup idea aimed at reducing food waste in Australian homes by leveraging the AI-driven development platform caffeine.ai.

The group’s concept is a virtual pantry app that helps users track their groceries and avoid duplicate purchases or forgotten produce. “With 2.5 million tonnes of food waste generated by Australian households each year, we knew our product was addressing a real, significant issue,” said one of the students, Jovia. The app is designed to provide a simple interface for users to log items, see what’s on hand, and receive alerts or suggestions to use what they have before it spoils.

Jovia highlighted the role played by caffeine.ai in accelerating their prototype development. “We used caffeine.ai to create our app’s prototype and were amazed at how easy it was to create. With a single prompt, our app was created; it was almost like caffeine.ai was reading our minds. It made the changes we wanted and listened to our prompts and instructions, to create the best app for our startup. We’re now hoping to represent our school at the final cross-campus pitch event. If selected, we plan to further enhance and polish our current prototype into a functional app, confident that caffeine.ai will help us get there.”

The startup is part of a school-wide project at Haileybury Berwick, where pupils were challenged to devise startup ideas ahead of a cross-campus event. Jovia’s group opted for the virtual pantry app concept, pitching its potential to deliver both environmental and practical household benefits. With supermarket purchases, shifting shopping behaviour, and mounting pressures on household budgets, technology solutions that reduce waste are receiving growing interest.

Caffeine.ai itself is built on the Internet Computer Protocol (ICP) and allows users to build and deploy applications through natural language prompts, bypassing much of the traditional coding workflow. Caffeine+3Chainwire+3Peerlist+3 The platform is currently live in alpha and providing free daily credits to help users experiment and prototype ideas.

For Jovia and her classmates, the appeal lies in the rapid iteration, accessibility and the fact that they could focus on the idea and design rather than programming details. The team plans to deliver a fully functional version of the app if selected for the final pitch. If successful, they aim to refine the prototype into a consumer-ready product.

There are several challenges ahead. The household food-waste problem in Australia is complex, involving consumer behaviour, supply-chain inefficiencies, perishability of produce, and the cost of waste disposal. According to recent figures, the 2.5 million tonne estimate represents a sizeable burden on sustainability goals and household budgets.

Implementing a virtual pantry app raises questions of user uptake and habit change. Will households consistently log their purchases? Will the system integrate with supermarket data or require manual input? How will food-expiry tracking and alert fatigue be handled? Additionally, the app’s utility will depend on its accessibility, ease of use and value proposition for everyday consumers.

From a technical angle, using caffeine.ai means that the core app logic may live on chain or at least benefit from a blockchain-based backend. That could improve transparency and resilience, but may also face challenges around scalability, cost and integration. Many startups built with no-code or low-code tools fail to bridge the gap between prototype and live product; the conversion from an AI-generated prototype to a production-grade, user-ready app remains work-intensive.

Still, for a high-school project the team is off to a promising start. The idea taps into a widely acknowledged problem, uses emerging tech in a hands-on way and offers learning outcomes for the students around design, entrepreneurship and sustainability. Jovia’s observation that the AI “was almost like … reading our minds” suggests the students were able to experiment rapidly, iterate ideas and focus on the user-experience rather than getting stuck in code. That kind of shift in educational settings is significant.

If Jovia’s team succeeds, the app could serve as a model for youth-led startup innovation paired with AI tools. And for those in the household food-waste space, it presents a fresh contender in the race to turn ideas into measurable change.

Find out more about the project on—Nexus and explore Caffeine to see how it works.