Back to portfolio
Fintech

The Market Maker

A contrarian thesis on credit access, from someone who built the models that price people out.

Illustrative archetype, not delivered work

Screen
Validate
Build
Match
Post-MVP
The problem

SME credit underwriting in Bangladesh and Thailand relies on collateral and formal financials — two things early-stage businesses structurally can't provide. The risk models are calibrated for the wrong population.

What we’re building

Alternative-data underwriting layer using cash-flow signals from mobile money, inventory turns, and supplier relationships — inputs the incumbent models ignore.

The founder

Spent 7 years at a regional bank building the credit scoring models that declined 60–70% of SME applications. She can articulate, in precise risk-model terms, exactly why those declines were conservative rather than accurate — and what data would change the output.

The problem

The market is not failing because lenders are unwilling. It is failing because the model is trained on data that excludes the population it's supposed to serve. That's a data problem with a data solution.

Why it cleared the screen

  • $50M+ SAM: conservative, defensible, bottom-up
  • Founder has direct access to the population (former colleagues at two banks as distribution partners)
  • AI-native advantage: the alternative-data pipeline is the moat; no manual underwriter can replicate it at the transaction speed required

This is an illustrative archetype representing the kind of founder Apon Venture Lab builds with. It is not a real case study or delivered work.

See yourself in this archetype?

Apply, or try a free resource to validate your idea first.

Try a resourceApply
Apply to Build With Us