Not a claim — the dataset. Every point is a real simulated deal, colored by how it actually came out and arranged by credit tier. Filter the cloud and watch the proof reorganize.
How this is built: The real dataset is 3,743,006 simulated deals. The canvas renders a faithful, proportional sample of them — each point colored by its real outcome (90.5% funded, 8.8% declined) and placed by its real credit tier (the bands from deep-subprime core to super-prime rim). Outcome and tier counts are pulled straight from the simulation database.
800,000+ deals of learned judgment, yours on your own server. One price, forever — and you can run a real one before you ever talk to us.