Nary LY

• Developed Various Machine Learning Models such as Propensity Models, Credit Risk Models, Look Alike Models based on customers' behavior patterns across various products that customers had utilized on the Wing Bank app to enhance customers’ experience, maximize profit, and reduce credit risk.
• Performed Vintage Analysis on customers’ app usage and transaction trends to track behavior trends of Wing customers over the years of operations to get insight and act on time to maximize profit based on trends finding.
• Derived and Enhanced variables from existing raw data to usable insight such as customer 360 likes variables for statistical modeling and rule-based projects.
• Create and maintain tableau dashboards to track the performance of digital loan products that were developed through Machine Learning models such as Propensity Models, look-alike models, and Credit Risk Models to see whether the ML models that were developed are still effective.

Intern as a data engineer for 3 months, and got promoted to full-time staff, worked on face recognition project, object detection, and building footprint.

• YoloV7 with TensorFlow was the main library to build the Building Footprint Deep Learning Model. This project was built to help the real estate company to evaluate land pricing.
• Object detection was built for a similar reason to the building footprint project, but this project evaluates the objects that exist in the house to evaluate the house price.
• The Face Recognition Project was built mainly for face scanning attendance to be used internally using Facenet.

Works as a desktop application developer Intern for an automated robot coffee arm. This project was mainly developed for the canteen but then the organization decided to not launch it due to the high cost of the materials. This project served as my second year thesis defense which got92/100 as evaluations from juries