Summary
Xiwen Wu is an algorithm engineer focused on multimodal AI and AGI pursuits, combining strong academic grounding (MSc in Computer Engineering from NUS, top-ranked BE in AI) with four years of hands-on experience in diffusion and video generation systems. He has published at an AI conference, interned on AIGC text-to-image diffusion models, and recently contributed to a multimodal LLM and production video-generation services at Baidu and HPC-AI Tech. His work spans model optimization, data engineering, and inference backend development—improving convergence, automating dataset curation, and scaling APIs to 10,000+ daily calls. Xiwen also explored PTQ for diffusion models and established benchmarks for anime image generation, showing a rare blend of research rigor and product-focused deployment. Notably, he combines sensitivity to data quality (aesthetics, motion, watermark filtering) with practical prompt refinement techniques that materially improve T2V/I2V/C2V outputs.
4 years of coding experience
1 year of employment as a software developer
Bachelor of Engineering - BE, Artificial Intelligence, Ranked 2nd in major, Bachelor of Engineering - BE, Artificial Intelligence, Ranked 2nd in major at Northwestern Polytechnical University
Master of Science - MS, Computer Engineering, 4.58/5.0, Master of Science - MS, Computer Engineering, 4.58/5.0 at National University of Singapore