Sheng Wang

Main Responsibilities

• Lead the development of internal tools and AI solutions to empower non-technical users with search analytics and operational insights.
• Provide end-to-end visibility on algorithm initiatives through dashboards, reporting, and A/B test result tracking for stakeholders.
• Drive assortment growth strategies through data-driven methods, including ranking, recall, and traffic optimization in search and recommendation.
• Act as the central point of contact between business teams and algorithm resources, translating business needs into structured, actionable roadmaps.
• Coordinate solution delivery across internal and external algorithm teams, ensuring alignment with strategic priorities and timelines.

Key Achievements:

• Built and launched a scalable traffic-boosting pipeline that significantly improved item visibility and conversion across multiple surfaces.
• Designed and deployed a GBDT-based model to identify potential hero assortments, paired with a three-stage traffic incubation framework to validate performance.
• Led the development of an internal LangGraph-based AI agent powered by GPT-4, enabling natural language relevance reviews, query rewriting, and SQL-free data access.
• Developed a query recommendation engine for search ads using embedding similarity between items and queries, improving targeting efficiency for sellers.
• Contributed to core search optimization projects, including:
  – Virtual Title generation to enhance recall
  – RankQ2Q for selecting optimal query rewrites
  – Search Side Filter to improve relevance of displayed filter attributes