Developed a recommendation system for bill promotions in the GoPay app, leveraging collaborative filtering to optimize CTR and conversion rates, achieving a 10-15% uplift over BAU. Optimized and maintained GoPay’s product recommendation section, conducting experiments to enhance CTR, ensuring above 5% daily CTR. Developed an end-to-end data science solution for GoPay Later acquisition, optimizing personalized voucher allocation using Uplift Modeling & Cost Prediction. Achieved a 25% conversion (GoPay Later activation) rate increase and 66% cost savings compared to the BAU approach.

Developed an ML classification model to detect illegal processed food sales on e-commerce to complete the undergraduate thesis in collaboration with BPOM. Scrapped and manually labeled 10K+ text data, experimented with Random Forest, KNN, Multinomial Naive Bayes, and XGBoost, utilizing Bag of Words and TF-IDF for feature extraction. Achieved a best F1-score of 0.859 using Random Forest and Bag of Words.

Guided high school students in developing a disease classification and chatbot project in NUS and SUTD. Mentored a team that secured 2nd place in the event’s hackathon.