Hdf: not at all necessary. As long as the machine has already analyzed an app from a broker or bank statements from the a borrowers bank, it doesn’t matter whether it’s the same app or not as long as it has the experience with that app and continues to improve through an automated feedback look — and through advanced ML, we are able to get to a much higher higher degree of accuracy than humans can and it continues to improve its accuracy by teaching itself, thereby improving every day with each incremental app. So let’s say we start with 50k apps and back statements from thousands (no idea the number there) banks; that’s already enough info to address most banks in the country (and with extreme accuracy for all the major or semi major banks given amt of data) accurately and Apps even more accurately (assuming data on app is right). As long as you feed your machines enough information and allow them to “solve a problem” instead of, as folks do it today, “solve for a specific solution” (sounds like a minor nuance, but it’s a major difference in machine learning), where the “best” in the mkt today are from a technology standpoint will still be outdated.

Call me Monday, bud.