Ads Core ML: Ranking & Foundational AI.

Unattributed data learning: Unattributed data has been underutilized in the Ads domain. Our current project aims to generate synthetic data using both rule-based and modeling approaches, then leverage these data to enhance a wide range of Ads foundational model, ranking model and retrieval models.

Area tech lead for Ads Deep Funnel Optimization, including Ads Conversion, Ads Lead Generation, Ads Qualified Lead and Talent Lead products. Recent projects focus on privacy-preserving conversion modeling, and value-based optimization.

Ads Targeting Optimization and ROI modeling for Linkedin Ads.

Ads CTR & CVR modeling and Target Optimization for Linkedin Ads.

Develop production-grade geophysical big-data processing applications (C++) in Compute Infra Team, R&D. Before this role, I also worked a year for business unit for seismic signal and image processing.