A young artificial intelligence startup with Filipino roots has raised $4.5 million in seed funding to tackle one of lending’s oldest problems: figuring out whether borrowers with limited traditional credit histories are actually creditworthy.
Kita, an AI-powered credit assessment company co-founded by Carmel Limcaoco and Rhea Malhotra, secured the fresh capital in a round led by BoxGroup, with participation from Y Combinator, Golden Gate Ventures, BEENEXT, Kaya Founders, U.S. News Digital Ventures, and other investors.
The financing puts another spotlight on a growing segment of financial technology where artificial intelligence is being used not simply to automate customer service, but to reshape the way lenders evaluate borrowers.
And Kita is targeting a particularly difficult part of that market: people and small businesses whose financial histories do not fit neatly inside conventional credit bureau databases.
Kita Wants AI to Read the Financial Records Banks Struggle With
Traditional underwriting often depends on standardized records such as credit scores, banking histories and formal financial statements.
But millions of potential borrowers—particularly in emerging markets—earn, spend and save money across a much messier collection of records.
Those can include bank statements, payslips, invoices, screenshots, scanned documents and electronic-wallet transaction histories.
Kita says its technology can analyze those documents, extract relevant financial information, identify inconsistencies and turn the resulting data into a structured credit assessment for lenders.
The company describes its platform as an AI-native underwriting system, rather than an autonomous lender.
That distinction matters.
Kita says its system can recommend an assessment and prepare underwriting materials, but the lender retains responsibility for the final credit decision.
Its current product stack includes document-reading technology, an AI-powered credit officer that can communicate with borrowers through channels such as WhatsApp, SMS and email, and an AI underwriter designed to prepare decision-ready credit memos.
The company’s Southeast Asia offering is specifically designed to process documents commonly encountered in the Philippines and Indonesia, including payslips, bank statements, e-wallet records and business documents.
More Than $130 Million in Loans Have Passed Through Kita’s Platform, Company Says
Kita says it has already processed more than $130 million in loan volume globally.
The company also says its technology is being used by microlenders, small and medium-sized enterprise lenders and commercial lenders in markets across Asia, Latin America and the United States.
Its website lists deployments in the Philippines, Indonesia, Mexico, South Africa and the United States, while Kita says more than 100,000 borrower files have passed through its technology.
The company has also claimed that underwriting workflows that could previously require extensive manual work can, in some cases, be completed in under 60 seconds using its platform.
Those figures are company-reported and should not be interpreted as independent measurements of lending performance or default outcomes.
Still, they help explain why investors are betting on the technology.
The larger opportunity is not simply making existing loans faster.
It is potentially allowing lenders to economically assess borrowers they might previously have ignored because reviewing their records manually was too expensive.
Why the Philippines Could Be an Important Market
The Philippines is particularly relevant to Kita’s strategy because a significant portion of financial activity increasingly happens outside traditional branch banking.
Digital wallets, fintech applications, informal businesses and alternative income sources have created enormous volumes of financial data—but not all of that information flows neatly into conventional credit scoring systems.
For lenders, that creates a paradox.
A borrower may have months or years of legitimate financial activity but still have what the lending industry calls a “thin file”—a limited traditional credit history from which to calculate risk.
Kita’s argument is that artificial intelligence can turn those scattered financial records into something an underwriter can actually use.
Its Southeast Asia platform is configured to interpret locally relevant documents and payment information, including records connected with platforms such as GCash and Maya, as well as payslips, bank statements and business filings.
Kita says it is already live with enterprise lenders in both the Philippines and Indonesia.
The Startup Began at Stanford
Kita’s story also carries a Silicon Valley connection.
Limcaoco and Malhotra started developing the company while studying computer science at Stanford University.
Limcaoco has described the company as evolving from an early university project into an underwriting platform serving lenders across several regions.
The startup has attracted backing from investors with experience in technology and fintech, including Y Combinator and Golden Gate Ventures.
The latest round was led by BoxGroup, an early-stage venture capital firm whose historical investments have included companies such as Stripe and Plaid.
Other participants reported in the financing include Genting Ventures and Apex Star Capital, the family office of Xiaomi co-founder Lin Bin, alongside several angel investors.
Philippine business executive Lisa Gokongwei-Cheng was also identified among the individual backers.
Where the $4.5 Million Will Go
Kita plans to use the new capital primarily to expand its engineering operation and strengthen its underwriting and fraud-detection technology.
The company is also preparing for additional demand in Southeast Asia, Latin America, Africa and the United States, according to reports surrounding the funding announcement.
That expansion could place Kita inside an increasingly competitive market.
Banks and fintech companies around the world are experimenting with machine learning, alternative data and generative AI to automate parts of the lending process.
The prize is substantial.
Reducing the cost of underwriting could allow financial institutions to profitably offer smaller loans and evaluate more borrowers.
But credit is also one of the areas where mistakes made by artificial intelligence can have serious consequences.
AI Can Make Lending Faster—but Can It Make It Fairer?
The real challenge for companies such as Kita may therefore come after the technology demonstrates that it can process documents quickly.
Credit decisions involve more than extracting numbers from bank statements.
Lenders must also determine whether the underlying models are accurate, explainable and free from unacceptable bias.
There are equally important questions about privacy.
Financial documents can contain some of an individual’s most sensitive information, making data security, storage and regulatory compliance critical as AI-powered lending systems become more widespread.
Kita says borrower data is processed within relevant geographic regions and that lenders maintain control over final decisions. The company also says it is pursuing SOC 2 Type II and ISO 27001 compliance work.
Those safeguards could become increasingly important as regulators around the world scrutinize how AI is used in consequential decisions involving money, employment, insurance and access to essential services.
The Bigger Opportunity Is the Borrower Banks Never Saw
Kita’s $4.5 million financing is modest compared with the enormous funding rounds associated with frontier AI companies.
But the business problem it is attacking could be much larger than the size of the round suggests.
If artificial intelligence can reliably interpret the fragmented financial histories of borrowers who lack conventional credit records, lenders could suddenly have a viable way to evaluate millions of people and small businesses previously considered too expensive to assess.
That could mean more loans, faster approvals and potentially wider financial inclusion.
It could also give financial institutions dramatically more information about borrowers than they have ever possessed.
Kita has already shown investors that there is demand for technology capable of converting messy financial documents into underwriting intelligence.
The harder question comes next:
Can AI help lenders see creditworthy borrowers that traditional systems missed—without introducing an entirely new generation of risks along the way?
Leave a Reply