Job Description:
I. Machine Learning Research & Development
- Develop and experiment with models, including deep learning ranking models, transformer-based architectures, and large-model-enhanced retrieval or reranking methods.
- Explore innovative approaches such as generative ranking, multi-task learning, sequence modeling, and vector-based retrieval.
- Conduct offline research using Shopee's large-scale datasets and evaluate model improvements in terms of AUC, NDCG, recall@K, and latency.
II. Applied Data Science & Analytics
- Perform deep-dive analysis on user search behavior, query intent, click-through patterns, and content features.
- Build data pipelines to process and validate large-scale logs using Spark, Hive, or PyArrow.
- Conduct A/B test analysis, interpret experiment results, and recommend further improvements.
- Identify root causes for search degradation, diagnose model blind spots, and propose data or feature improvements.
III. Production Support
- Collaborate with software engineers to deploy models into Shopee's multi-stage search architecture (retrieval ranking post-ranking).
- Implement efficient inference pipelines and monitor model performance in production.
- Optimize models for large-scale production constraints such as latency, memory, and throughput.
Requirements:
- Strong foundation in machine learning, deep learning, or information retrieval.
- Proficiency in Python and experience with ML frameworks such as PyTorch or TensorFlow.
- Solid understanding of data structures, algorithms, and linear algebra.
- Experience working with large datasets and distributed data processing tools (e.g., Spark).
- Ability to independently structure experiments, analyze results, and draw actionable insights.
Preferred Qualifications
- Research or practical experience in ranking models, transformers, session-based recommendation, or vector search.
- Hands-on experience with ANN libraries (e.g., FAISS, HNSW, ScaNN), graph algorithms (e.g., Swing, SSG, NSG), or generative recommendation systems.
- Understanding of large-scale system constraints such as memory-efficient models, quantization, or serving optimization.
- Familiarity with SQL, feature engineering pipelines, or search system components (query understanding, intent prediction, content relevance).
- Strong communication and collaboration skills ability to work with cross-functional product and engineering teams.
What You Will Gain
- Exposure to real-world search and recommendation system challenges at massive scale.
- Opportunities to explore cutting-edge research areas including generative ranking, agent-enhanced search, and multi-modal retrieval.
- Experience contributing to high-impact production systems used by millions of users daily.
- Mentorship from experienced scientists and engineers in one of Southeast Asia's leading e-commerce companies.
- Potential pathways to full-time roles in machine learning, data science, relevance engineering, or applied research.
- Potential research papers on applied data science track for top-tier ML or Data conferences.
