GISAID
The world's largest COVID-19 database, searchable by the people who curate it.
Sep 1, 2022- Client
- GISAID
- Services
- Internal tooling
- Year
- 2022
- SECTOR
- Health tech
01 — The ChallengeWhat we were handed
Pulling data the world would act on, from sources it could trust.
The hard part was never the volume. It was building functionality that could pull from trustable sources and then let the scientists who use it correct what it found — so a record that is wrong does not stay wrong while everyone waits for a release.
In the hand
The tool, on the desk it runs on
02 — How it runsTech & architecture
Seven algorithms doing the matching, against twenty million records.
MatchingNLP and sequence matching with BioPython, across seven algorithms
SignalsDate of record, type of disease, author records and origin of virus
At scaleBLAST, run against the twenty million records GISAID holds
SurfaceA dashboard built for the internal team rather than for a demo
03 — Outcomes
A tool their own researchers run, on the data they already trusted.
7Matching algorithms behind one comparison
20MRecords the matcher runs against
Start hereTell us what you are building. We will come back with what it takes and what it costs.
Start a project like thisNext caseThe President’s ClubA Samsung thank-you, engineered as a membership app.
Open case