Our recent insights

AI can create real value in lending. It can also create token cost, governance burden, integration friction, and explainability problems when used where rules, code, or analytics would do better.

Lending teams should stop asking “Can we use AI here?” and start with the workflow problem, decision consequence, and best tool for the job.

Most lending stacks have improved data access and decisioning tools, but outcomes still lag. Carrington Labs’ Chief Executive Officer (CEO) and Chief Product and Commercial Officer explain what’s missing—and how to close the “donut hole” with decision-ready analytics.

Monitoring is operational risk control. Learn what to track for drift and stability, how to set review cadence, and how to run change control that stands up to governance.

Test length should be driven by event rates, decision volume, and your outcome window. This guide shows how to size lending experiments so results are decision-grade and governable.