GScore: Unlocking Financial Inclusion for Filipinos

Session Outline

A look into GCash’s journey to enable millions of unbanked Filipinos access to credit, using ML models trained on transaction and other data sources data to create the PH’s first trust score: GScore. This session will focus on sharing the story of how this score came about, from its purpose to provide cheaper credit for the unbanked, the key ML components that helped streamline its execution, and its quantitative impact to the GCash base.

Key Takeaways

• Key Point 1: We saw that unbanked Filipinos needed access to cheaper credit
  There are a lot of Difficult barriers to entry for Filipinos to access formal credit and predatory lending was pervasive. We saw an opportunity in leveraging our rich transaction and other data sources to create an alternative credit score using machine learning for a more inclusive credit access.

 

• Key Point 2: We created GScore, the country’s first trust score
  A primer on GScore and the key ML components that helped enable its streamlined execution.

 

• Key Point 3: GSCore now has the ability to provide the right credit amount to X millions of Filipinos to fit their needs and lifestyle.
  Quantitative metrics and qualitative stories showing the impact of GScore to the GCash base.

 

• Key Learnings / Takeaways
  ● Promote the value of empowering “Fair access to credit, for all”
  ● Invest in collecting the right data
  ● Invest in data engineering and infrastructure

————————————————————————————————————————————————————

Bio

Iñigo Benavidez – Head, Data Science & AI GCash Data Office | GCash | Philippines

For almost six years, he has been leading teams of data scientists, solving a diverse range of AI problems, from credit scoring, fraud detection, to recommender systems. He and his team pioneered GScore, the country’s first trust score, which has enabled millions of Filipinos access to credit, with tens of billions of pesos disbursed since launch. Today, Inigo is looking to further expand the impact of AI at GCash, both through researching and developing new AI use cases at his company, as well as hiring exceptional talent to help him and his team discover those use cases. Beyond work, he loves diving into books, almost to a fault — a typical weekend involves finishing one but buying three others. His favorite genres are sci-fi and non-fiction.

 

————————————————————————————————————————————————————

Bio

Mico Comia – Lead Modeler for GScore Machine Learning Research Data Science & AI GCash Data Office | GCash | Philippines

Mico is a rookie Data Scientist — being introduced to the field in 2019 during his undergraduate thesis, when he worked on a model that aims to understand the risk factors that lead to student attrition at the UP Diliman College of Engineering. The study was awarded Best Paper at the 2020 International Conference on Engineering Education. Shortly after graduating with a degree in BS Computer Engineering, he joined GCash’s inaugural FinTech Cadetship Program. He then transitioned to his role as a Machine Learning Engineer, assisting the GScore team and creating an in-house scorecard modeling module to transition out of a sunsetting platform the team used. Currently, he is taking up his MSc in Computer Science and is a Lead Machine Learning Researcher for GScore — creating its latest version, which simplifies the architecture to address scalability and explainability. In his free time, he likes listening to music and watching musicals.