Collecting data is the start. Making it usable is the work.
How we go from thousands of disconnected records to a product someone relies on every week.
Why it is hard
The information an industry runs on is spread across agencies at every level of government and across the private sector: permits, licenses, registrations, inspections, filings and more, changing every day. No two of those systems were designed to work together.
Each one names things its own way, stores them in its own format and changes that format when it needs to. The same company appears under different spellings. The same address is written three ways. A record that mattered yesterday is replaced today with no notice.
The information is there. What is missing is the work of making it agree with itself.
That work is slow, repetitive and never finished, which is exactly why so few people do it well. It is the work we chose.
How we work
We watch. We monitor government agencies and private-sector sources on infrastructure we own, continuously, so changes are picked up as they happen rather than months later.
We keep the original. Every record is stored exactly as it was published before anything else happens to it. Whatever we conclude later can always be checked against what the source actually said.
We resolve. We work out which records describe the same company, place or project, and connect them. Once that is settled, it stays settled, and everything built afterward benefits from it.
We build. Only then do we build a product, and we keep it narrow: one question, one kind of customer, answered clearly enough to act on.
What you can expect from anything we publish.
Traceable
Every value can be followed back to the source it came from.
Current
Sources are checked continuously. How quickly a record appears depends on how quickly the agency publishes it.
Fixable
If something is wrong, tell us. We check it against the original record and correct it at the source of the error.
Common questions.
Where does your data come from?
From government agencies and private-sector sources, collected by acquisition systems we build and run on our own infrastructure. The collection, and the history it builds up over time, are ours.
How current is it?
Sources are checked on a loop measured in hours. A record reaches our products as soon as practical after the agency publishes it.
What happens when a source changes its format?
The engine is built to notice when a source changes shape, adapt to it and keep running. Changes like that are the most common reason data feeds fail quietly, so we treat them as routine rather than exceptional.
How do you decide what to build next?
We look for industries where the records already exist, the people who need them cannot practically use them, and a focused product can prove itself quickly. We announce products when they launch.
I think something about my business is wrong. What do I do?
Email [email protected] with what you are seeing and where. We will check it against the original record and follow up.
Want to go deeper?
If you work with messy data and want to compare notes, we would like to hear from you.