Zeyi Cheng is a CEO and engineering leader with nine years of experience specializing in inference performance and production ML systems, now based in Monaco. He founded WaveSpeedAI after driving GPU-accelerated deep learning deployments at Alibaba, where he integrated 50+ models, optimized TorchScript engines, and achieved 1.5x gains over PyTorch’s cuDNN path while processing millions of documents cost-effectively. Prior roles include inference engineering at SiliconFlow and service and payment optimization at ByteDance, reflecting strong full-stack and systems-level chops. His GitHub focus on building the fastest inference optimization framework for generative AI underscores a rare blend of low-level performance tuning and large-scale product impact. Zeyi is migrating critical pipelines to PyTorch 2.0 and has delivered record-breaking Stable Diffusion improvements for image editing and inpainting. He combines operator-level rigor with founder ambition, prioritizing efficiency where GPU cost directly affects product margins.
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.