I am engaged in the development of the Payment Foundation Model, where I design Large Language Models to analyze billions of transaction records. The objective is to derive deep insights into patterns for various downstream tasks, including the prediction of transaction life cycles and fraud detection. Additionally, I am working on improving the effectiveness and efficiency of model architectures by enhancing Transformer and State Space Models, such as proposing linear self-attention mechanisms and designing non-causal Mamba for non-sequential features.

I was responsible for the development of rigorous formal methods for verification of leaning-enabled cyber-physical systems such as autonomous driving, as well as the design of provably safe machine learning components such as deep neural networks.

I was responsible for the development of the application program interface and softwaredevelopment kit for a smart sensor that fuses inertial measurement and visual tracking.

I was responsible for designing a salary management system based on the Visual Basic environment.