The new Income Prediction Model from Credolab requires no personally identifiable information, financial history, or documents, enabling lenders to assess creditworthiness even when traditional income data is unavailable.
Credolab, a behavioral and device metadata analytics company, has launched its new Income Prediction Model, a machine-learning solution that estimates income levels using anonymized smartphone metadata.
As financial institutions increasingly encounter applicants with limited or informal income records, access to reliable income data remains a key challenge, particularly in emerging markets. Credolab’s new model addresses this gap by analyzing anonymized signals such as app usage patterns, device model and age, premium VPN usage, and other behavioral indicators captured through its lightweight SDK.
The system is fully privacy-compliant, relying only on data collected with explicit user consent and never accessing age, gender, education, contacts, messages, or other personal content.
The model works by transforming raw smartphone metadata into millions of engineered behavioral features and identifying the most predictive patterns. For example, usage of investment, luxury retail, airline loyalty, and premium shopping apps often correlates with higher income, while payday-loan, gambling, and discount-shopping apps are more common among lower-income users.
The age-adjusted value of a user’s device, as well as premium VPN usage, also provides strong signals of earning power. By training the model on verified income samples from the client’s customer base, it achieves high accuracy across diverse populations, including thin-file, informal-income, and new-to-credit users.
The Income Prediction Model is currently rolling out across Latin America, with Brazil and Mexico as initial focus markets, and will soon expand into Southeast Asia. Leading banks and fintechs are already integrating the solution to improve underwriting for thin-file and informal-income customers, helping extend fair credit to underserved populations. Each deployment is tailored using verified income data from the client’s customer base, ensuring accuracy and alignment with local population characteristics.
“In many markets, a lack of verified income data is the biggest barrier to financial inclusion,” said Peter Barcak, CEO and co-founder of Credolab. “Our new model gives lenders a privacy-safe and statistically sound way to infer income levels using only device behavior. It’s a powerful step toward fairer, faster, and more inclusive credit decisions, especially among populations for whom traditional data simply doesn’t exist.”
Credolab’s Income Prediction Model reinforces the company’s mission to expand access to credit while mitigating risk, fraud, and reliance on incomplete financial data, offering banks and fintechs a powerful tool for the evolving lending landscape.
