Point Predictive today launched Predictor Agents, a Deep Intelligence AI fraud tool that gives lenders three new ways to manage fraud before funding and during servicing, covering applications they receive and, for auto lenders, dealers who send them. Predictor Agents draw on data from more than 650 consortium lenders, including more than 340 million applications and $6.0 trillion in risk-assessed loans.
Losses that hurt most often end up in news. Dealer pushing falsified paystubs through dozens of deals, ring of synthetic identities building credit for bust-out, or straw borrower scheme spread across several lenders can cost millions and draw attention from regulators and reporters.
Predictor Agents review every consortium application in real time for identity, income, employment, dealer risk, straw borrowing, credit washing and bust-out activity. When they find applications from single lender or from multiple consortium members that should be unrelated but tied together by common thread such as phone number, employer, address or dealer, they flag pattern so lender can act before money goes out.
Fraud that ends up in headline usually leaves trail months earlier. Predictor Agents show underwriter that trail while pattern still developing, not after losses pile up.
Scores and alerts describe one application at time. Predictor Agents give risk and fraud teams direct access to query data repository intelligence covering application histories, loan performance, negative files and fraud patterns from 650+ lenders, along with own contributed loans. Teams can ask questions about whole portfolio in plain language and get answers immediately.
Chief lending officer can use Predictor Agents to tune underwriting strategies and rules to better target early payment default risk. For example, executive can ask which parts of book producing loans that default in first six to 12 months, then tighten rules on those segments without slowing good borrowers.
Auto lenders can also research dealer activity in depth to make more informed decisions. Analyst can ask how dealer's applications perform across entire consortium and whether same employers, phone numbers or addresses keep appearing, then decide whether to add reviews, change terms or grow relationship. Benchmarking built in as well. Credit union can compare its early payment default and fraud rates against similar cohort of consortium lenders to see where loans perform better or worse than market.
Full manual forensic review of one loan file can take experienced analyst 20 minutes to hour. Agentic Forensic Fraud Reviews from Predictor Agents produce same kind of report in seconds, giving small fraud team reach of much larger one.
Behind each review are Predictor Agents, specialists trained by Point Predictive fraud team on trillions in risk-assessed loans. Each agent applies what team has learned to its own area such as identity, income and employment, dealer risk, straw borrowing, credit washing and bust-out activity. Every report lays out agents findings by review area, supporting evidence and what to do next. If file needs closer look, report spells out what to verify before funding, and if file checks out, explains why loan can move ahead.
About Point Predictive
Point Predictive powers new level of lending confidence and speed through artificial intelligence, powerful data insight from proprietary data repository, and decades of risk management expertise. Company's data and technology solutions quickly and accurately identify truthful and untruthful disclosures on loan applications. As result, lenders can fund majority of loans without requiring onerous documentation, such as paystubs, utility bills, or bank statements, improving funding rates while reducing early payment default losses.
