Shelf Friend helps my friend compare supermarket prices and product care
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built A friend needs help while shopping at a supermarket: compare prices, understand how a kitchen or household item should be used, and weigh its advantages against the maintenance it brings home. I built Shelf Friend , a small Chinese-language web app with two connected decisions: what costs less per usable…
This story is about Shelf Friend, a small web application designed to help a friend compare supermarket prices and understand proper product usage. The app is a Chinese-language tool that uses the MiniLM open-weight multilingual model to compare prices and provide relevant usage guides. When comparing items, the app takes into account the net quantity and pack count, ensuring accurate calculations.
The AI then retrieves a guide or buying checklist from a small corpus of relevant content. While a cheaper item may not always be the best choice, Shelf Friend helps users make informed decisions. The app's code is publicly available under an MIT license, allowing for inspection and improvement. To build Shelf Friend, the developer utilized Transformers.js and the quantized ONNX conversion of the multilingual MiniLM model.
The app demonstrates the power of open innovation, as the semantic search functionality runs within the user's browser, eliminating the need for a remote inference service.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
