Run it on your own book. Keep the findings.
The pilot is a retrospective fraud exposure audit on your real history. It is free, and the findings are yours regardless of what you decide next. Nothing leaves your building.
Three ways to check the work, agreed before we run.
Success criteria are set with you up front, then measured against your own data. You decide whether it earned a second look.
Backtest against known losses
We replay the signals against accounts you already charged off, with targets agreed before the run. Because the replay is point-in-time, the lead time it reports is what you would have seen in real time. We report the result as measured, whatever it is.
Re-review of closed alerts
Take the accounts DoubleCheck flags that your team already closed as false positives. Re-open a handful and judge for yourself whether the cultivation pattern is real.
The thirty-minute readout
A short readout of the memo, ending on one question: over the last few years, how many of your charge-offs were members who were never real.
If the signals miss, we report that as is.
This is a measurement, not a sales demo. If the backtest is weak on your book, the memo says so in plain terms. A pilot that finds little is a real result you get to keep. We would rather tell you the truth than win a logo.
About two afternoons of your team's time.
One afternoon
Run the extraction script against a core export and hand off the de-identified files. It is a single file with a self-test.
One afternoon
Walk the flagged accounts and the re-review set. Confirm what looks real against what you know.
Thirty minutes
The readout and the one question. Decide whether it is worth continuing.
Four weeks, start to readout.
Scope and extract
Agree the account scope and success targets. Your IT runs the extraction script.
Replay
The engine runs on your machine. Signals replay point-in-time across the history.
Review
Your analyst walks the flagged accounts and the re-review set against known outcomes.
Readout
The memo and the thirty-minute readout. You keep the findings.
Pilot is free. Production pricing is discussed at the readout, once you have seen the findings on your own data. Request a review →