Liquidity
Insufficient-funds behavior and near-term solvency stress.
- Overdrafts
- Overdraft protection transactions
- ACH returns
- Returned checks
- Transaction reversals
Cash flow based lending reads the account itself, and grades it against the lines your policy draws.



Delinquency starts as an account running short, debt the income cannot carry, or risky spending. We catch it while the file is still an application.
Insufficient-funds behavior and near-term solvency stress.
Capacity to carry a debt load relative to income.
Repayment consistency and adherence to scheduled obligations.
Discretionary or high-risk behavior correlated with credit stress.
Non-traditional income that still carries a repayment.
Account reliability and whether this is the primary bank.
No missed payment, no collection, nothing to report yet. The strain shows in the account first, and it is graded the moment it does.
An account that averaged less than one overdraft a month just had four in one. The month-over-month read shows the momentum, not just the count.
Every finding carries its date, so a file that went wrong last month is never read like one that went wrong last year.
Each flag's condition maps the reading to a level — nine overdrafts in six months lands this file in Warning — and the file carries its worst finding, set by your policy, not ours.
A flag is two things you own: whether it runs at all, and the condition it trips on. Both are set per program, so a near prime book and a subprime book never share a cutoff.
Every check in the catalog is switched on or off by you, and a signal you already rely on is computed here with the standard set, so one read covers your policy rather than most of it.
Every flag ships with a default condition and level, and they are replaced with yours. The same overdraft count can be clean on one program and a Warning on the next.
Looking for the number these flags sit beside?
KoraScoreWant the whole file this read sits on?
Cash Flow AnalysisWorried the document itself is fake?
Prevent fraudWhat has to hold before a cash-flow signal is allowed near a credit decision.
Every flag tells you exactly what triggered it and what level it’s set at, so you can read the rule yourself instead of just taking our word for it. It’s never a black box, and it’s never us making a judgment call on your behalf.
No, a flag just means a condition from your own policy got hit, at the level you set. What you do with that is up to you: price the deal, ask for more documentation, or decline it.
It still shows up as a zero, it doesn’t just disappear on you. That matters, because “nothing happened” and “we never actually checked” are two very different things. And if a rule depends on some other check being turned on, it’ll say so in its own condition.
Every flag starts out with a default rule and cutoff, but you can swap in your own for each program so your near-prime book and your subprime book don’t have to share a cutoff they shouldn’t. Most teams check back in every quarter, or whenever their risk appetite changes.
Yes, any time new information comes in. We don’t overwrite anything; every run sits on its own, so if someone asks what you knew when you made a call six months ago, you can actually go back and check.
Send an applicant you have already decided. We read the twelve months behind them, run the flags at the lines your program sets, and show you the level each one resolves to.