Summary
Ju-chiang Wang is a Research Scientist and Tech Lead based in California with a Ph.D. in Electrical and Computer Engineering and nine years of industry experience bridging music research and production-grade AI. He has led deployment of 20+ deep learning models into products like TikTok, CapCut, and Ripple, and co-developed award-winning audio separation models (Mel-RoFormer/BS-RoFormer) that won the Sound Demixing Challenge 2023 and power voice separation features in major apps. His work applies advanced Transformer architectures to music information retrieval tasks—achieving state-of-the-art results in transcription, event detection, and music generation—while also contributing to recommendation, auto-tagging, and mashup/remix systems. Proficient in Python and C++ with a strong algorithmic foundation, he has published 50+ papers and consistently moved research into scalable product features. Less obvious: he blends academic rigor with product-first engineering, acting as the point of contact to shepherd research prototypes through full deployment at scale.
9 years of coding experience