Work / Global trade · 2017–2022
Deciding whether a trade was real
The product financed trades between a buyer and a supplier who had usually never worked together, across borders. Almost nothing about either party had been written down anywhere a stranger could read it.
Before money moved, the program had to answer whether the transaction was real and whether the parties were who they claimed to be. Beneficial ownership, sanctions and PEP screening, adverse media and physical verification, run across every entity a transaction touched, with behavioral signals from the verification team’s own app alongside them.
What came out of it was a standard profile of how a counterparty actually performs, which the market adopted as its own way of reading a supplier.
Screening on every transaction before funding, and the profile adopted as shared vocabulary across the market
What’s in it
- What the decision actually was
- Most parties were exactly what they appeared to be, which was the problem
- What it had to catch
- What know your business meant here
- The signals that came back from a visit
- Every entity, and the relationships between them
- The output was a decision, not a number
- The profile that came out of it
- What it taught