Shivi Sharma dedicated ten years to working in credit risk roles at companies such as American Express and Varo Bank.
Over time, she noticed that teams were devoting the same effort to evaluating every loan, whether it was for $100,000 or $5 million. This meant that processing smaller loans became a time-intensive and unprofitable task for lenders.
Together with her husband, Utsav Shah, she saw a chance to address this issue.
“She observed that most small business owners struggled to secure the funding they needed to expand, simply because the economics didn’t add up for banks,” Shah explained to TechCrunch.
“With our combined experience in building large-scale AI-driven decision systems and our background in credit and fraud risk management within financial services, we realized we could leverage advanced AI agent workflows to tackle this long-standing challenge,” he added.
The couple went on to establish Kaaj in 2024, a startup focused on automating credit risk evaluations so that underwriting can be completed in minutes instead of days. Kaaj reports having processed over $5 billion in loan applications, serving clients like Amur Equipment Finance and Fundr. On Wednesday, the company revealed it had raised $3.8 million in seed funding from Kindred Ventures and Better Tomorrow Ventures.
Here’s how the platform operates: When a small business applies for a loan and submits all required paperwork (such as financial records, bank statements, and tax documents), underwriters would traditionally spend several days manually checking and entering this data into their Loan Origination System (LOS).
Kaaj employs AI to extract, categorize, authenticate, and organize this information within the LOS. It also performs checks for document alterations for the underwriter fraud team. The system integrates with popular Customer Relationship Management (CRM) platforms like Salesforce, HubSpot, and Microsoft, and can even indicate whether a business meets a lender’s policy requirements.
“This enables a team that previously handled 500 applications a month to process 20,000 with the same number of employees, making smaller loans financially feasible,” said Shah, who serves as CEO.
The goal is to make it easier for small businesses to obtain loans from banks by reducing the cost and effort required for banks to evaluate them.
Other companies in this space include Middesk, Ocrolus, and MoneyThumb. Sharma believes Kaaj can differentiate itself by automating the full credit analysis workflow, not just segments of it.
“We achieve this by implementing agentic AI workflows that replicate the work of underwriting teams, supporting lenders in reviewing complete loan packages,” she explained.
The newly raised funds will be used to speed up product innovation and extend their reach among independent and small business lenders. “We’re committed to advancing our AI agent technology, broadening our module selection, and growing our network of lenders and brokers beyond our current scope.”
Ultimately, Shah and Sharma hope that Kaaj will help transform small business lending by bringing automation to a process that still relies heavily on paperwork.
“By automating the technical side of credit analysis, we allow human underwriters to concentrate on the nuanced aspects of deal-making and subjective evaluation, which is where their real strengths lie,” he said.
