• Implemented the build system changes to support the organization-wide effort to migrate the development of the Android platform to the trunk-based development with feature flags.

• Migrated the partitions packaging rules from Make to Soong, achieving 30% reduction in clean and incremental analysis time.

• Focused on migrating the build system of the Android platform from Soong, Ninja and Make to Bazel.

• Simplified the dependency chain of the Java API stub generation, improving the clean and incremental build performance of the Android framework by 30%, significantly enhancing the productivity of the platform developers.

• Developed and simulated reinforcement learning robot simulation environment implementing MuJoCo engine and OpenAI Gym.

• Explored feasibility of reinforcement learning adaptation in factory manufacturing environment, modelling multiple agents based on algorithms such as Proximal Policy Optimization and Soft Actor Critic.

• Certified Advanced Programmer for excellence in efficient programming and sophisticated knowledge in algorithms and data structures.

• Enhanced better understanding of data structures in C++ of students by answering questions and guiding debugging during daily office hours.

• Promoted active discussion of students during weekly lab sessions, exploring various data structures including trees, hash table and heaps.

• Conducted and assisted signal processing and data analysis based on the vital signals collected from the integrated sensors system.

• Engineered algorithm development of the company product's key features, including posture detection and fall detection, implementing Support Vector Machine classifier.

• Developed a web-based interactive dashboard implementing Flask, providing interdepartmental assistance based on insights discovered from data analysis and signal processing of sensor-accumulated patients' vital signals.

• Created and trained a digit-recognition Convolutional Neural Network model with custom datasets using PyTorch, significantly reducing the difficulty and increasing the accuracy in gold data extraction.

• Led and developed a GUI program development project using PyQt, reducing the clinical researcher’s hardship in gold data extraction by minimizing manual labelling.

• Gained knowledge in digital signal processing, including frequency filtering, peak detection and transform.