Decision Engine
Decision engines automate MCA underwriting decisions through rule-based logic and machine learning models — enabling instant or near-instant approvals on standard submissions while routing complex deals to human underwriter review.
Why This Matters
Decision engine architecture: ingests submission data (bank statements, application, credit, third-party data), applies funder-specific rules (minimum revenue, time-in-business, credit thresholds), runs ML models (default prediction, optimal advance amount, optimal pricing), and outputs decision (approve/decline/refer to human review). Top-tier engines deliver decisions in 60-300 seconds for 70-80% of submissions, with remaining 20-30% routed to human underwriting for nuanced judgment. Decision engine quality directly affects funder competitiveness — fast decisioning wins broker preference and merchant conversion.
Frequently Asked Questions
Frequently Asked Questions
What share of MCA underwriting is now automated?
60-80% at sophisticated funders for standard advances; near 100% for renewals on existing merchants; lower percentages for larger advances ($250K+) and complex business types. Automation share continuing to increase.
Do decision engines outperform human underwriters?
On standard submissions yes — engines apply consistent criteria without fatigue or bias. Humans outperform on edge cases requiring nuanced judgment (unusual business models, recovering merchants, complex business structures). Best operations combine both.