FINTECH

Building a platform that reads M-Pesa statements in seconds

How we turned a slow, manual credit assessment process into a web-based analytics platform now used by multiple lenders across Kenya.

Amua 2023 4 min read
Seconds
Analysis time
M-Pesa & Bank
Statement types
Live product
Status

"What used to take an analyst an hour per application now takes seconds. And the consistency is something we could never achieve manually."

Lending Operations Manager

In Kenya, M-Pesa is more than a payment method. For millions of people, it is their primary financial account. When someone applies for a loan, their M-Pesa statement is often the most complete picture of their financial life available to a lender. But reading that picture used to take hours.

The challenge

Lenders assessing creditworthiness need to understand income patterns, spending behaviour, and risk signals from transaction history. Analysts were doing this manually, scrolling through thousands of transactions, categorising them by hand, and building mental models of each applicant’s financial profile. The process was slow, inconsistent, and hard to scale.

Bank statements added another layer of complexity. Different banks use different formats, transaction descriptions vary wildly, and the sheer volume of data in a six-month statement makes manual review impractical for high-volume lending.

What we built

Amua is a web-based platform that parses M-Pesa and bank statements automatically. It categorises transactions, identifies income sources, maps spending patterns, and generates a borrower profile with key financial indicators. What used to take an analyst hours now takes the platform seconds.

The system handles the messiness of real-world transaction data. It recognises common transaction patterns, deals with inconsistent formatting, and surfaces the signals that matter most for credit decisions. Lenders get a clear, standardised view of each applicant regardless of which bank or mobile money provider they use.

From client project to product

Amua started as a solution built for a single client. Once we saw how well it worked, and how universal the problem was, we turned it into a standalone product. Today, multiple lenders across Kenya use Amua to process loan applications faster and more consistently.

This is a pattern we see often at Revolution Analytics: a real problem encountered while working with a client becomes the foundation for a product that serves the broader market. Amua is proof that the best products come from solving problems you have seen first-hand.

Working on something similar?

Get in Touch