Summary
Soon Soon is a pragmatic software engineer with 12 years of experience building high-performance back-end systems, ML compiler backends, and large-scale data pipelines across companies like NVIDIA, Groq, Twitter, and Amazon. He combines deep systems and compiler expertise in C++ and Rust with practical ML and production tooling (PyTorch, TensorFlow, MLIR, Kafka) to move research prototypes into reliable, cost-saving production services. His work has driven measurable operational impact, including a six-figure annual cost reduction at Twitter and design contributions to S3 control plane scalability at AWS. A University of Toronto MASc graduate in recommender systems, he blends rigorous research (SIGIR publication) with hands-on engineering—often favoring working prototypes over slide decks. Colleagues will find him detail-oriented, thesis-trained in graph embeddings, and fluent across languages from Verilog and assembly to Python and Scala.
12 years of coding experience
9 years of employment as a software developer
Computer Engineering Artificial Intelligence, Computer Engineering Artificial Intelligence at University of Toronto
c/c++, python, java (programming language), English, Chinese, Chinese