AI-based Algorithm to Crunch 1Million Features from Opt-in Smartphone Metadata to Generate Customised Digital Credit Scores

Abstract/Technology Overview

We use smartphone device data to generate a behavioral score for the customer that predicts her probability of defaulting. By accessing anonymised, metadata of the device, our AI-powered algorithm sends the financial institution the score of the applicant in a matter of seconds. In doing so, we not only open a whole new segment of customers for banks and lenders, but we also bring down their cost of risk, speed up the processing time, and increase their overall Gini Coefficient. The algorithm takes into account over 1 million features before arriving at the score that predicts the creditworthiness of applicants. To heed to the different nature of customers in different markets and industries, our data scientists work with our clients to create an algorithm that best suits their businesses.

Technology Features, Specifications and Advantages

Our unique competitive advantages include:

  1. A scalable and replicable business model in any market with no credit bureau data or that in segments, such as the millennials, self-employed, micro-SMEs, underserved and unbanked.
  2. Seasoned and experienced risk, product, data science and technology specialists. The only alternative credit scoring company founded and led by a career Chief Risk Officer.
  3. Our algorithm has been built and perfected from more than 8 million loan applicants across 49+ lending partners. It crunches nearly 1 million features from opt-in smartphone metadata to find the most predictive delinquent behavioral patterns before converting them into credit scores - all this in a matter of seconds.
  4. Highest possible GINI coefficient or Kolmogorov–Smirnov (KS) statistic with AI-powered, tailor-made scorecards developed on the client's anonymous credit outcome data. Building and improving the algorithm constantly by taking inputs from clients using our solution and in compliance with the regulators whom we are in direct contact with. This gives us an upper hand in being the first movers in the alternative data-based credit scoring space.

Potential Applications

Over two billion people are unbanked today. Many of them are creditworthy but lack the credit history needed by financial institutions to approve their requests for a loan. The narrow, conventional definition of creditworthiness, and the limited parameters used by most financial institutions to make credit decisions, usually leads to rejection. As a result, they miss out on opportunities to build or grow their business, or to get credit for any of their personal needs.

We are on a mission to help financial institutions make better credit decisions. We do this by redefining the way banks view creditworthiness and by giving them access to, perhaps, the best source of behavioral data in the world: mobile device data. By leveraging our superior alternative source of data, we give financial institutions the power to underwrite applications from unbanked, new-to-bank, and new-to-credit such as the millennials, self-employed, and thin-files. We do this with high predictability, great accuracy, full privacy protection, and in real-time.

Contact Person

Michele Tucci


CredoLab Pte Ltd

Technology Category
  • Artificial Intelligence
  • Financial Technology
Technology Status
  • Available for Licensing
Technology Readiness Level
  • TRL 9