Overview
Nutry pulls scattered nutritional data into one place and makes it instantly searchable. Instead of hunting across labels and databases, you search a food, see its aggregated nutritional breakdown, and compare it against others — a fast, focused tool for understanding what's actually in the things we eat.
What it does
- Instant food search — fuzzy, typo-tolerant search across a large aggregated food dataset, powered by Elasticsearch so results come back as you type.
- Aggregated nutritional breakdowns — macro and micronutrient profiles normalised across sources into a single, consistent view per food.
- Compare & analyse — line up foods side by side to spot differences at a glance, with analytics over the underlying dataset.
- Accounts & saved data — sign-in and per-user state via Firebase.
How it's built
A single-page React app with Redux + Redux-Sagas managing data flow and the asynchronous search/analytics pipelines. The search layer is backed by Elasticsearch, which does the heavy lifting for fuzzy matching and ranking over the aggregated corpus. Firebase handles authentication and lightweight persistence, the app is hosted on AWS, and the pipeline is wired up with Jenkins for CI and Jest for the test suite.
The interesting engineering was less the UI and more the data side: ingesting and reconciling nutritional information from different sources into one normalised schema that Elasticsearch could index and serve quickly.