- PIC and Mentorship: In charge of below services and collaborate with team members. For high traffic services, I take up the duty to stand-by during campaign hours. - User Targeting Service: A service justifies if an user meets the configured displaying requirements, like GPS location and user labels. It serves high customer-facing traffic and requires collaboration with many upstream and downstream services. - User Targeting SDK: Based on original service logic above, we made a SDK version to reduce network and server cost. - Image Processing Service: A tool service semi-auto generates promotion images, by C++ image processing libraries and connecting with AI team services. It’s a general processing system which is low-code to support new AI modules. - URL Redirection Service: URL redirection is highly related to FE implementations, which worked with FE and Nginx logics closely.

- Banner Scheduling: A set of microservices displays promotion images on Shopee website and app which serves high daily traffic. - System Refactorization: We migrated the backend system from Python mono repository into Golang microservices. As for admin system, we refactored the logics to make it low-code supporting new incoming requirements.

Developed the backend server and the Portal website with the team members.