Market Analysis
Brazil's Credit Market: Alternative Data's Role
Brazil has one of the most sophisticated credit bureaus in emerging markets, yet 35 million adults remain without a credit file. Alternative data is the bridge.
Insights
Credit intelligence, research, and insight
Practical analysis: alternative credit data, financial inclusion economics, and AI-led underwriting for lenders in Brazil and Latin America.
Ethics and Compliance
Responsible AI in Credit Underwriting: What It Really Means
Using machine learning to score borrowers raises real questions about fairness, explainability, and accountability. Here is how to navigate them.
Business Case
Financial Inclusion ROI for Lenders
Lending to borrowers without formal credit histories is not charity. For lenders with the right scoring infrastructure, it is a growth strategy with measurable returns.
Lending Strategy
How Lenders Reach First-Time Borrowers
First-time borrowers are the segment most lenders avoid, but avoiding them is not the same as managing the risk. Here is how smarter scoring changes the calculus.
Data Science
Thin-File Borrowers and Machine Learning: Effective Approaches
Classic ML models trained on bureau data perform poorly on thin-file populations. Feature engineering with alternative signals changes what is possible.
Data Science
Utility Payments as a Credit Risk Signal
Electricity and water bills are paid before rent in most low-income households. That payment discipline is a strong signal, if you know how to read it.
Industry Perspective
Why Traditional Credit Scores Miss Half the Market
The bureau system was designed for people with bank accounts, credit cards, and formal employment. It was never designed for the majority of the world's population.
Data Science
Prepaid Mobile Activity as a Credit Signal
How someone manages their prepaid mobile account, when they recharge, how much, how consistently, turns out to be a reliable proxy for how they manage other financial obligations.
Data Science
Credit Modeling Without Bureau Data
When your target borrower has no credit file, the standard modeling toolkit does not apply. Here is how to think about feature selection, validation, and calibration from scratch.
Market Analysis
Alternative Data for Mobile First Credit Scoring
Mobile penetration in Latin America now exceeds formal banking by a wide margin. The data trail from mobile activity is already richer than the bureau data most lenders rely on.
Industry Perspective
Latin America's Unbanked Credit Gap
An estimated 200 million adults in Latin America lack access to formal credit. This is not a social failure. It is a data problem, and data problems have engineering solutions.