The decision and the limit come out together.
Score, income and debt-to-income, with the history already in your database. Approve, decline, or flag for review with the reason — and the amount you can release comes from the same table that decided, not from another system.
From application to released amount
Bureau, open finance, your own data and, when it makes sense, a document read by the pipeline.
Synchronous: fetching from the sources dominates the time, and the decision comes back in the same response.
With the policy version frozen in the decision record.

What the rules usually read
None of this is fixed: these are the inputs your policy declares. They are the ones that show up in almost every consumer credit policy.
- Score
- The bureau's, yours, or both. If you already compute an in-house score, it comes in as one more field, and the policy decides on it alongside the market one.
- Income
- Declared, estimated by the bureau or confirmed by a document — and which one counts is a rule, not a habit.
- Debt-to-income
- How much of the income is already committed to debt. The most tweaked threshold in any policy.
- Relationship
- Tenure, late payments and product usage — what's in your database and no bureau knows.
- Request
- Amount and term requested. The rule can approve on its own up to an amount and flag for review above it.
What sits in the middle comes out flagged
A good score with high debt-to-income, income that doesn't match the amount requested, a name match on a watchlist: the rules flag it for review, and the decision keeps everything the machine saw.