Kita Is Automating Credit Review in Markets Where Banking APIs Do Not Exist

The Macro: Billions of People Have No Credit Score

In the United States, applying for a loan means pulling a FICO score, running it through an algorithm, and getting an answer in minutes. In the Philippines, Indonesia, Vietnam, and dozens of other emerging markets, that infrastructure does not exist. Most of the population is unbanked. Banking APIs are nonexistent. A borrower’s financial history lives in e-wallet screenshots, bank statement PDFs, utility bills, and handwritten records.

Credit and risk teams at lenders in these markets do everything manually. An analyst opens a PDF, reads through transaction histories, types numbers into a spreadsheet, and makes a judgment call. This process takes hours per application and caps how many loans a lender can underwrite in a day.

The fintech boom in Southeast Asia has created hundreds of digital lenders, but most of them are bottlenecked by this same manual review process. Open banking standards that would solve this are years away in most countries. Plaid does not work here. Finicity does not work here. The data exists, but it is trapped in documents.

Kita, backed by Y Combinator, builds document intelligence specifically for lending in these markets. They extract, structure, and analyze financial data from the messy document formats that emerging market borrowers actually use.

The Micro: Reading the Documents No One Else Can

Carmel Limcaoco (CEO) and Rhea Malhotra (CTO) cofounded Kita with deep knowledge of both the fintech and emerging market sides of this problem. Based in San Francisco but focused on Southeast Asian lending markets, they are building the document intelligence layer that these markets desperately need.

The product handles the full pipeline: ingest documents in various formats (PDFs, screenshots, scanned images), extract structured data (transaction amounts, dates, balances, income patterns), and produce a credit-relevant analysis that risk teams can act on. The key technical challenge is handling the diversity of document formats across different e-wallets, banks, and utilities in each country.

In the Philippines alone, borrowers might submit GCash e-wallet screenshots, BDO bank statements, Meralco utility bills, and GrabPay transaction histories. Each has a different format. Each changes its layout periodically. A system that works today might break when GCash updates its app design next month.

Competitors in the broader document intelligence space include Hyperscience, Rossum, and various OCR providers. But general-purpose document AI struggles with the specific formats and languages used in Southeast Asian financial documents. Kita’s vertical focus on emerging market lending gives them domain expertise that horizontal players lack.

The business model is clean: lenders pay per document processed, and the ROI is immediate because it replaces manual labor. If an analyst spends 30 minutes reviewing documents for one application, and Kita does it in seconds, the math speaks for itself.

The Verdict

Kita is solving a real, painful problem for a large and growing market. The emerging market lending boom is real, and document processing is the genuine bottleneck.

At 30 days: how many lenders are actively processing applications through Kita, and what is the error rate compared to manual review?

At 60 days: can Kita handle new document formats without significant retraining, or does each new e-wallet or bank require custom work?

At 90 days: are they expanding beyond the Philippines into other Southeast Asian markets? The playbook should transfer, but each country has its own document formats.

I think Kita is building exactly the right product. The emerging market credit infrastructure gap is not going to be solved by open banking standards anytime soon. Document intelligence is the practical bridge, and Kita is building it.