• Lead developer and associate designer of VR therapeutic application "ADAPT_XR", which aims to provide a comprehensive online or blended physical therapy solution using VR/AR technology for physical therapists and related patients.• Led the design of application framework, program architecture, and flowchart. - Developed two demonstrative demos for partner evaluation. - Actively participated in project development planning (e.g., development cycle and sprint tasks) from summer 2023 to spring 2024.• Led the development of the foundational framework components, including object pools, singletons, and scriptable asset managers, ensuring reusability and efficiency for future development.• Designed a general-purpose state machine and standardized its framework and APIs. Based on the framework, developed two state machine variants - animation state machine and system state machine - for handling animations and managing project functionalities and systems.• Contributors:- Ahmad Alsaleem: project manager, designer.- Roger Alan Altizer: designer- Ziheng Mo: lead developer, associate designer- Hongke Wei: artist- Anirudh Dargar: developer

• Main developer of mobile application “Beat Blade” (Based on C# and Unity, Android and iOS platform).• Developed highly reusable UI system, guidance system and asset management system using singleton design pattern, C# async, C# reflection, and Unity scriptable assets, which reduced maintenance complexity and accelerated development timeline.• Developed Unity UI plugin “DynamicScrollRect” using object pool and C# delegate. DSR reduces runtime memory cost by 75% and UI culling cost significantly compared to Unity's ScrollView component, and supports user-customized layouts.• Developed a customizable state machine framework for handling system states, animations and other development scenarios• Integrated analytics SDKs including Firebase, Facebook SDK, Kochava, IronSource and developed tools for unified API and functionalities management, enabling codeless customization of data collection strategies for producers.

• Individual contributor of Ascend cloud computing platform,a C/C++-based platform that integrates machine learning (ML) frameworks including TensorFlow, PyTorch, and MindSpore. Ascend offers a full-stack solution for AI research and development.• Developed ML operators (L1LossGrad, Ger, Slogdet, etc.) using Da-Vinci architecture, which allows neural networks to compute in the unit of vectors and matrices.- Achieving over 30% increase in performance compared to ARM and CUDA architectures.• Transpiled and integrated "LAPACK" and "BLAS" libraries into Ascend system, streamlining ML operator development pipeline.- Authored two technical documents for library integration and operator development guidance.

• Associate developer of IKEA's BESTA Indoor Visualization System (Based on C/C++ and AprilTag, Windows platform).• Enhanced AprilTag algorithm for BESTA, enabling webcam-based object detection, improving resistance to environmental light interference, and optimizing code base for Raspberry Pi deployment. - Reduced production costs by 50%.

• Actively participated in the development of the open-world-based RPG game "Newbie At CUHK(SZ)" (Based on Unity, Windows and MacOS platforms) from April/2020 to July/2020• Actively participated in the research of the natural language processing project "Plot Generation Based on Crowd Intelligence" (Based on Tensorflow and Python) from June/2019 to April/2020

Worked as a research associate fellow in with Atomic, Molecular and Optical Physics research group in "High Energy Laser Lab" in University of Sussex. The title of project is "Investigating Non-linear Optical Loop Mirrors"