HomeSelected engagements

05 / 05Case study

Salary and recurring inflow detection

Kuda Technologies · African neobank, 7M+ customers

Salary and recurring income read straight from raw transactions, including accounts the supplier could not cover.

35%

coverage uplift over the third-party supplier

Income detection · Kuda Technologies

+35%

salary and recurring-income coverage over the third-party supplier

no data no data no data no data
Supplier only most accounts come back with no data
no data no data
Detection framework finds the regular salary in raw transactions

Illustration of the method. Coverage up 35% over the supplier, and up 45% on accounts where it returned nothing, at a 9% false positive rate.

Built with Python, SQL, Algorithmic design

01

Problem and solution.

The problem

Lending affordability assessment required reliable detection of salary and recurring income. Third-party supplier (Mono) had limited coverage, particularly on accounts with sparse transaction history.

What I built

Designed an algorithmic framework for salary and recurring inflow detection from raw transaction data, replacing reliance on a third-party supplier with limited coverage. System directly feeds into credit eligibility and affordability decisioning.

  • Python
  • SQL
  • Algorithmic design

02

Architecture.

Schematic

  1. 01 Raw transaction data
  2. 02 Inflow detection framework (detection step) salary and recurring inflows · algorithmic replaces third-party supplier (Mono)
  3. 03 Credit decisioning
    • Affordability
    • Credit eligibility
Schematic. Component names from the engagement summary.

03

Results.

  • 35%

    coverage uplift over the third-party supplier

  • 45%

    uplift where the supplier returned no data

  • 9%

    false positive rate

  • Directly improved credit eligibility accuracy for underbanked customers.