Data Scientist

AI/ML Development Mandaluyong City, National Capital Region, Philippines

Profile
Contacts

Summary

At Grab, I led the development of a machine learning pipeline to detect multi-app driversthose using competing platformswhich directly affected our supply reliability. Faced with a highly imbalanced dataset (700 out of 60,000), I engineered features like 8-week behavioral variance to capture pre-churn signals that traditional volume-based metrics missed. I used kNN for oversampling and XGBoost for classification, achieving an AUC of 0.75 and enabling precise driver segmentation. With this, I automated incentive logic that dynamically adjusted based on predicted behavior, increasing loyal driver earnings by 5 percentage points and reducing incentive setup time from 3 days to 2 hours. I built the entire pipeline using Python, SQL, and Databricks, from data extraction to real-time dashboardingowning both model performance and business impact end-to-end.

Expectations

Im looking for a role where I can apply data science to solve real business problemswhere the work goes beyond modeling and directly influences strategy, operations, or product. I want to be part of a team that values both creativity and structure, where I can contribute to building frameworks, automating decisions, and scaling solutions that make a measurable impact. Its important to me that I work in a culture that encourages open thinking, fast execution, and continuous learningsomewhere I can grow not just technically, but also as a strategic partner to the business.