The small business loan application is a mess of documents. Bank statements, tax returns, Secretary of State filings, and a dozen other forms pile up, waiting for an underwriter to spend days piecing together a story. Kaaj AI says it can do that job in under three minutes. Founded in 2024, the San Francisco startup is selling an agentic AI platform that ingests the entire document package and spits out a decision-ready credit memo, claiming to power thousands of these loans each day [Finovate, Fall 2025].
The Underwriting OS Wedge
Kaaj’s product is built as a workflow engine that plugs into a lender’s existing loan origination system. It automates the tedious, document-heavy analysis that precedes the final credit decision. The platform’s sequence is methodical: it ingests and classifies over 100 document types, performs KYB checks against public records, analyzes cash flow from bank statements, and runs a fraud detection scan inspecting 25 forensic signals in under five seconds [Kaaj.ai, 2024]. The final output is a lender-specific credit memo, formatted to that institution’s internal policy.
A Team Built for Scale and Risk
Founders Utsav Shah and Shivi Sharma built their careers at the intersection of high-scale AI and credit risk. Shah spent a decade at Uber and Cruise building AI-powered decision systems. Sharma is an expert in credit and fraud risk from stints at American Express, Uber, and Varo Bank [Kaaj.ai, 2024].
| Founder | Role | Key Background |
|---|---|---|
| Utsav Shah | Co-Founder & CEO | AI decision systems at Uber and Cruise |
| Shivi Sharma | Co-Founder & President | Credit and fraud risk at American Express, Uber, Varo Bank |
The SMB Lending Tailwind
Kaaj is launching into a market defined by both massive scale and persistent friction. The US small business loan market is valued at over $1.4 trillion, and new business formation has surged, with a record 430,000 new applications filed monthly in 2024 [Kaaj.ai, 2025]. Kaaj’s automation pitch targets this inefficiency, aiming to help banks, credit unions, and non-bank lenders scale their SMB operations. The company is listed in the American Bankers Association partner network [American Bankers Association, 2026].
The Execution Hurdles
For all its promise, Kaaj’s path is not without obstacles. The competitive landscape includes established players like Ocrolus for document processing and Middesk for business verification. Kaaj’s differentiation rests on bundling these capabilities into a single, agentic workflow, but convincing risk-averse lenders to adopt a new, AI-driven core process is a steep sales cycle. Success will hinge on moving from generic claims to named, referenceable deployments.
Kaaj’s $3.8 million seed round, led by Kindred Ventures with participation from Better Tomorrow Ventures, Karman Ventures, Pythia Ventures, and Coughdrop Capital, provides the runway to tackle these challenges [Kindred Ventures, 2024]. The question for the next twelve months is whether Kaaj can convert that pedigree and early capital into a shortlist of named enterprise lenders.
Sources
- [Kaaj.ai, 2024] Home and About Us pages | https://kaaj.ai/
- [Finovate, Fall 2025] FinovateFall 2025 profile | https://finovate.com/videos/finovatefall-2025-kaaj-ai/
- [Kindred Ventures, 2024] Investment announcement | https://kindredventures.com/announcement/kaaj-building-the-intelligence-layer-for-small-business-lending/
- [Better Tomorrow Ventures, 2024] Why We Invested in Kaaj | https://better-tomorrow-ventures.ghost.io/why-we-invested-in-kaaj/
- [Kaaj.ai, 2025] Market sizing blog post | https://kaaj.ai/blog/seed-round-funding-announcement
- [American Bankers Association, 2026] Partner Network directory | https://www.aba.com/experts-peers/partner-network/directory/kaaj-ai