KredosAI Raises $7M to Scale Behavioral Collections

KredosAI closed a $7 million Series A led by BMW i Ventures, with new investors Motley Fool Ventures and Walter Ventures joining existing backers Okapi Venture Capital, StartFast Ventures, SaaS Ventures, and Stout Street Capital, according to BusinessWire. The round was oversubscribed. GeekWire reports the Issaquah, Washington company was founded in 2021 by former T-Mobile executives Balaji Sridharan and Dave Thoms, and that total funding to date is now a little over $10 million. BusinessWire reports the new capital will fund three priorities: expanding into financial services and auto lending, building out a multi-agent framework with voice-agent support, and growing the team 2x to 2.5x over the next 12 months. CEO Sridharan told GeekWire the company's premise is that "the majority of consumers who go late on payment actually want to pay."

BusinessWire and GeekWire attribute performance figures to KredosAI itself: the company says its platform has cut write-off rates by 11.5% and lifted customer lifetime value by 13.6% across its enterprise portfolio, worth more than $50 million in annual bottom-line benefit for some Fortune 50 clients. GeekWire reports KredosAI has processed more than 200 million customer interactions over two years, with revenue growing more than sixfold in that span, and that the company is SOC 2 Type II certified. These figures are company-reported and have not been independently verified by a named enterprise customer.

GeekWire reports KredosAI is a FICO partner, integrating its engagement layer into the FICO Platform used for underwriting and collections decisions, and that subprime auto-loan delinquencies have climbed to their highest levels since the 1990s, which BMW i Ventures partner Baris Guzel cited as a reason for backing the company as it expands into auto lending.

Enterprise behavioral-collections systems typically combine longitudinal account histories, channel-mix optimization, and treatment-effect estimation. Building a comparable system generally requires uplift modeling or A/B testing to isolate incremental repayment from baseline behavior, low-latency pipelines for near-real-time outreach decisions, and explainability tooling to support compliance review. Collections data is also heavily class-imbalanced and outcome-censored, since payments can arrive long after a default flag, which complicates standard classification approaches and pushes teams toward survival-analysis or time-to-event framings.

Whether independent case studies or named Fortune 50 references corroborate the 11.5%/13.6% figures, since all performance claims currently trace back to the company itself, and whether KredosAI's push into auto lending draws additional regulatory scrutiny of personalized debt-collection outreach.