Search & Recommendation Data Scientist(Intelligence Planning)
Core
Define search/recommendation problems and establish optimal algorithms and evaluation systems by analyzing global user behavior patterns and business context.
Role type
Senior IC data scientist (search & recommendation)
Builds
Search and recommendation platforms, evaluation protocols, and model specifications for global user experiences.
Domain
Search, recommendation, and content platforms
Deliverable
production ML models
Required skills
Statistical modeling, hypothesis testing, A/B test design, feature engineering, data requirement definition, model specification, evaluation protocol design
Preferred skills
Ranking/recommendation model optimization, localization strategy, causal inference, uplift modeling, embedding space analysis, semantic search, RAG/LLM integration
Technologies
Python, R, SQL, Pandas, Spark
Responsibilities
Analyze global user behavior data, search logs, and recommendation logs to define search/recommendation problems and improvement opportunities; Design statistical/ML model directions and prototypes for personalized recommendation, search ranking, user targeting, conversion prediction, and segment classification; Identify core features and domain variables for retrieval, ranking, recommendation, and RAG/LLM search; Propose application strategies for global search/recommendation algorithms considering regional user behavior differences and data bias; Establish offline/online evaluation metrics, experiment design, and A/B test-based evaluation systems for new algorithms or model candidates; Structure analysis results into model specifications, feature requirements, and evaluation protocols to connect with development teams for implementation and validation.
