Credit risk automation platform Kaaj raises $3.8M seed from Kindred Ventures

Shivi Sharma spent a decennary moving successful recognition hazard astatine places similar American Express and Varo Bank.  

At immoderate point, she realized teams were spending adjacent amounts of clip analyzing each types of loans — careless of whether it was worthy $100,000 oregon $5 cardinal — meaning assessing smaller loans was ultimately an unprofitable and time-consuming process for lenders.  

She and her husband, Utsav Shah, realized determination was an accidental here.  

“She watched as the immense bulk of small concern owners couldn’t access the superior they needed to grow, simply due to the fact that the economics didn’t work for banks,” Shah told TechCrunch.  

“Between our skills successful gathering AI-powered decision-making systems astatine standard and our expertise successful recognition hazard and fraud hazard assessments successful banking successful fiscal services, we realized we could use next-gen AI cause workflows to lick this decades-old problem,” helium continued. 

The joined mates decided to motorboat Kaaj in 2024, a institution that helps automate credit hazard investigation truthful that underwriting nary longer takes days, but minutes. Kaaj said it’s processed much than $5 billion worth of loan applications, with clients including Amur Equipment Finance and Fundr. The institution announced connected Wednesday a $3.8 cardinal effect circular from Kindred Ventures and Better Tomorrow Ventures.  

The merchandise works similar this: A tiny concern applies for a loan, submitting all the needed documents (like fiscal statements, slope statements, and taxation returns) — typically, erstwhile this happens, underwriters walk days manually verifying each this accusation and logging it into their Loan Origination System (LOS).  

Kaaj uses AI to identify, classify, verify, and organize accusation into LOS. It besides runs assessments to cheque for papers tampering for the underwriter fraud team. It integrates into existing Customer Relationship Management (CRM) systems similar Salesforce, HubSpot, or Microsoft and even shows a lender if a concern is gathering the criteria of a lender’s policy.  

“This allows a team processing 500 applications monthly to grip 20,000 applications with the aforesaid staff, making smaller loans economically viable,” Shah, the company’s CEO, said.  

The anticipation is that much small businesses volition beryllium capable to access loans from banks due to the fact that it becomes more cost-efficient for a slope to investigate them.  

Others successful the marketplace include Middesk, Ocrolus, and MoneyThumb. Sharma hopes that Kaaj will basal retired from the contention by automating the full recognition analysis process rather than parts of it.  

“We bash this by deploying agentic AI workflows that mimic their teams, to assistance lenders analyse end-to-end indebtedness packages,” she said.  

The caller superior volition beryllium utilized to assistance accelerate merchandise improvement and grow crossed autarkic and tiny concern lenders. “We’re focused connected enhancing our AI cause capabilities, expanding our module offerings, and scaling our lawsuit basal of lenders and brokers beyond our existent footprint.”  

Overall, Shah and Sharma hope Kaaj tin successful immoderate ways “revolutionize” tiny concern lending, bringing automation to what is inactive a precise paper-heavy process.  

“By automating the subject of recognition analysis, we free up human underwriters to absorption connected the creation of deal-making and subjective assessment, which is their existent competitory advantage,” helium said.