Principal Applied Scientist
Core
Design and implement ranking, reranking, and retrieval models using deep learning and LLMs for large-scale content recommendation systems.
Role type
Principal Applied Scientist (Recommendation Systems)
Builds
Next-generation ranking, reranking, and retrieval systems including generative recommendations and agentic feeds.
Domain
Internet / Recommendation Systems / Deep Learning / LLMs
Deliverable
production ML models (via careerplan.io/jobs/1970393556862646-principal-applied-scientist-at-microsoft)
Required skills
Deep learning, LLMs, recommendation systems, ranking models, search relevance, personalization, multi-task learning, contextual bandits, reinforcement learning, feature engineering, model training, evaluation, online inference, distributed pipelines, high-throughput online services, data structures, algorithms, asynchronous programming, large-scale data analytics, multi-objective optimization, PyTorch, TensorFlow, Spark
Preferred skills
Publications in top-tier ML/AI conferences, experience with agentic AI, generative AI applied to recommendation
Technologies
PyTorch, TensorFlow, Spark
Responsibilities
Architect ranking and retrieval systems, lead ML/DL model pipelines, establish technical standards for experimentation and model governance, partner with engineering and product teams, mentor team members, communicate team progress to leadership.
Seniority
Principal, strategy & mentorship
