Summary
Paul Xiong is a seasoned software leader and founder with 14 years of experience building applied AI and embedded systems, currently leading the open-source Cervical AI initiative from San Ramon, California. He combines deep hands-on engineering—spanning lightweight vision models for iOS/macOS, AutoML, and embedded Linux—with product-first startup instincts honed over decades as a CTO. His teams have shipped commercial AI camera and router products, and he has driven projects that processed large-scale video streams, ReID/face recognition, and mesh Wi‑Fi optimizations. Motivated by real-world health impact, he pivoted to cervical cancer screening AI, achieving 98% recall/precision on a dataset of 9,000+ patient images while tackling hard problems like overlapping-cell segmentation and extreme class imbalance. Paul prefers small, high-velocity teams of 5–6 engineers and is more energized by coding and product building than by traditional executive duties. He brings rare cross-domain experience—from kernel and network drivers to mobile vision models—paired with a pragmatic, research-informed approach to shipping life-saving ML systems.
15 years of coding experience
23 years of employment as a software developer
B.S, Computer science & Engineering, B.S, Computer science & Engineering at Beihang University