Summary
Shang-yu Su is a Principal Research Scientist based in Tokyo with 10 years’ experience building production-grade NLP, dialogue, and search/recommendation systems, holding a PhD in Computer Science from National Taiwan University. At Rakuten he moved search from 0→1, shipping multilingual semantic and multi-modal visual search that drove substantial GMS and CVR uplifts and delivered multimillion-yen annual cost savings. He designs end-to-end ML platforms—data pipelines, ONNX deployment, MLflow tracking, and human-in-the-loop LLM workflows—achieving large metric gains (100% NDCG@10 lift) and infrastructure savings via knowledge distillation. A hands-on leader, he has led cross-regional R&D teams and mentored interns while publishing industry work (SIGIR 2026) on LLM-based attribute generation. Notably, he blends rigorous research with pragmatic engineering, often optimizing throughput and experiment iteration speed by orders of magnitude to accelerate impact.
10 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at National Taiwan University
English, Japanese, Chinese, Korean