Nutrition Data and Scoring Infrastructure
Digital Nutrition Intelligence
The layer underneath both products: data processing at volume, and a scoring system that has to produce the same answer twice.
What it is
Digital Nutrition Intelligence is the shared data and scoring layer supporting our nutrition technology.
The system processes and organises data covering more than 800,000 verified food products, for use across both consumer and business applications.
The engineering problem
Food data arrives incomplete, inconsistent and duplicated. A product exists under several names, units disagree, and fields that matter are missing. None of that is visible to a user, and all of it determines whether the output is worth anything.
So the work is ingestion, normalisation, deduplication, verification and the scoring layer on top, plus the ability to reprocess when the methodology changes without losing the audit trail of what changed and why.
Scoring is deterministic by design. It is the difference between a number a dietitian can defend and a number that merely sounds right.
Why it is relevant to client work
Most companies with a data problem have this problem: the data exists, and it is not in a state anything can be built on.
The work of getting from raw sources to something dependable is the unglamorous majority of any data project, and it is the part that decides whether the dashboard on top means anything.
Capabilities demonstrated
What this project proves we can do
- Data ingestion and normalisation at volume
- Deduplication and verification across inconsistent sources
- Deterministic scoring with a reprocessing path
- A shared layer serving two products without forking
Technology
- ETL
- PostgreSQL
- BigQuery
- Python
- Scoring engine
Related services
Want this kind of work done for you?
More work
Other software we have built
Tell us what you are trying to build
Tell us about the business, the problem, what you have today, and what you want to accomplish.