Methodology
How we work with public data.
Our approach is simple: use public records, be clear about what we did to them, and keep every finding traceable.
- 01Public dataOfficial and public sources
- 02Clean & organizeNormalize records, resolve duplicates
- 03Detect signalsFlag patterns worth a look
- 04Deliver useful intelligenceClear reports with source references
1. Public data
We use official and otherwise publicly available sources, such as FDA public datasets and local health-department inspection records. We do not use private, purchased, or restricted data.
2. Clean and organize
Public records are messy. We standardize names and dates, remove duplicates, and match records that refer to the same device or business.
3. Detect signals
We apply defined rules to flag things worth attention, such as repeat findings or changes in volume. The rules are described in plain language so you can see why something was flagged.
4. Deliver intelligence
Reports are organized for a person to read and decide from. Where automation is used, a human reviews the output before it is treated as final. We are still developing these workflows, and we will say so plainly rather than overstate what is automated.
Limits
Public data can be delayed, incomplete, or inconsistent, and a record is not a conclusion. Our reports are informational and are not legal, regulatory, or compliance advice.
Questions about our method?
Ask us anything about sources or processing, or request a sample to see the output.