Rhea Malhotra, a Stanford-trained computer scientist and robotics researcher, is applying her artificial intelligence (AI) expertise to financial technology through Kita, a startup developing tools to automate loan underwriting and credit assessment.
Malhotra, who received Stanford University’s 2025 Firestone Medal for Excellence in Research for her work on robot learning in unfamiliar environments, is the co-founder and chief technology officer of Kita. The company uses AI to help lenders process borrower documents, identify potential fraud and prepare credit assessments.
In a LinkedIn post, Malhotra described how her Stanford research explored whether robots could perform tasks in environments they had never encountered. In one project, researchers trained a robot to open cabinets and put objects away, then tested its ability to do the same with unfamiliar furniture.
READ: Open for hiring: Kita raises $4.5 million to expand AI credit assessment for lenders (August 24, 2026)
She drew a parallel between that challenge and the problems lenders face when reviewing Small Business Administration (SBA) loan applications, where borrower documents vary and requirements change.
“How can AI help lenders put together an accurate, complete package for credit review under current SBA requirements?” Malhotra wrote, describing the issue her company is addressing.
From Stanford robotics research to financial technology
Malhotra completed her bachelor’s and master’s degrees in computer science at Stanford University, with a minor in mechanical engineering during her undergraduate studies.
Her research focused on robotics, mobile manipulation, and AI-powered machines’ ability to generalize what they learn to unfamiliar environments.
Her honors thesis, titled “Robust Robot Learning in Unseen and Dynamic Environments,” earned the 2025 Firestone Medal, Stanford’s recognition for outstanding undergraduate research. The university listed Malhotra among the recipients of its 2025 senior project awards.
At Stanford’s Interactive Robot Perception Learning Lab, she worked on research involving mobile manipulation, task generalization and robot learning. Her work included developing approaches that enable robots to perform household tasks despite changes in object appearances, layouts and surroundings.
One of her research projects, HoMeR, combines imitation learning with whole-body control to help mobile robots perform everyday manipulation tasks. The research demonstrated how robots could learn from a relatively small number of demonstrations and adapt to different household environments.
Malhotra also worked on humanoid robotics at Stanford’s Khatib Lab, where she contributed to research on human-like motion generation and robot movement.
Building AI tools for lenders through Kita
Malhotra co-founded Kita with Carmel Limcaoco, who serves as the company’s chief executive officer. The startup focuses on automating document-heavy lending processes, particularly in markets where borrowers’ financial records are scattered across bank statements, payslips, utility bills and other documents.
According to the company’s description on Y Combinator, Kita uses vision-language models to extract information from financial documents, verify data and generate signals that lenders can use for credit assessment.
The platform is designed to reduce the manual work involved in collecting borrower information, checking documents and preparing underwriting reports.
READ: Stanford signs Indian American guard Amisha Ramlall for 2027 (November 22, 2026)
Kita says its technology can help lenders identify inconsistencies and potential fraud while converting unstructured financial records into information that can support lending decisions. The company’s approach is particularly aimed at emerging markets, where traditional credit histories and banking data may be limited.
The startup was part of Y Combinator’s Winter 2026 batch and was founded in 2025.
AI playbook for SBA lenders
Malhotra and Limcaoco are scheduled to demonstrate the company’s approach to US small-business lending during a virtual session titled “Silicon Valley’s AI Playbook for SBA Lenders.”
The session, hosted by Kita, is scheduled for November 23 from 2 p.m. to 2:30 p.m. Eastern Time.
The demonstration will focus on how AI can support SBA 7(a) and 504 loan processing, including borrower document collection, financial analysis and automated credit memo generation.
The event is designed for SBA lenders, underwriters and credit operations teams. It will include a live demonstration and a question-and-answer session.
Kita says its technology aims to reduce manual bottlenecks, limit repeated requests for borrower documentation and help lenders accelerate credit assessments without expanding their teams.
Malhotra’s transition from robotics research to financial technology illustrates how AI techniques developed for machines operating in unfamiliar physical environments can also be applied to complex business processes involving variable documents and changing requirements.


