Made contribution to WasmEdge, an open-source WebAssembly (WASM) runtime. Implemented an advanced WASM thread proposal, introducing parallel processing to WebAssembly. This allowed for an improved thread scalability by 4.86x on 8 cores, surpassing multi-worker NodeJS, which reached only 3.74x.

Worked in the Sidewalk team under Ring, Amazon, concentrating on embedded android systems for intelligent devices. Integrated LLVM/LeakSanitizer into Android for Sidewalk. The introduced solution was more efficient over the traditional Valgrind, including 2x speedup, reduced memory usage, and a smaller binary size.

Researched on training large models on GPUs with limited memory capacities. Designed a cross-framework solution, achieving 2x speedup compared to NVIDIA's Unified Memory. The result were published to ASPDAC 2022.

Handled teaching assistant duties for Discrete Mathematics, Parallel Computing, and Advanced UNIX Programming. Assisted numerous students, collaborating closely with the instructor to design midterm examination questions.