Principal Applied Scientist, Advertiser Demand Intelligence

Microsoft Microsoft · Big Tech · Redmond, WA +1 · Applied Sciences

The Principal Applied Scientist will lead the development of ML and LLM components for a large-scale Demand Intelligence platform within Microsoft Advertising. This role involves defining scientific vision, architecting end-to-end ML pipelines, incorporating LLM-based features for narrative generation (including RAG and fine-tuning), and ensuring robust evaluation and Responsible AI compliance. The goal is to transform complex data into actionable insights, automate analysis, and streamline workflows, ultimately empowering stakeholders with smarter, faster decisions.

What you'd actually do

  1. Define and drive the modeling strategy for the advertising recommendations platform, spanning classical machine learning (for analytics on structured data) and the use of generative AI. You will set the direction on which problems to tackle with ML (e.g. predictive modeling, anomaly detection, clustering) and how to leverage LLMs to maximize user understanding and value.
  2. Architect end-to-end machine learning pipelines – oversee the design of data processing workflows, feature stores, model training/validation routines, and deployment mechanisms that can reliably produce daily insights for all customers. Ensure these pipelines are scalable, efficient, and maintainable, working closely with data engineering leaders on implementation.
  3. Lead the incorporation of LLM-based components for the platform’s intelligent narrative generation. This includes guiding the development of prompt frameworks, fine-tuning strategies, and retrieval-augmented techniques so that the system can answer complex sales questions and explain insights in conversational language.
  4. Oversee cross-team initiatives and collaboration, coordinating with engineering, program management, and stakeholder teams. You will chair technical design reviews, balance priorities, and guarantee that the data science efforts align with product requirements and timelines.
  5. Mentor and develop the applied science team, providing technical guidance to other scientists and engineers. Champion best practices in experimentation, coding, and MLOps, and foster a culture of scientific excellence and continuous learning.

Skills

Required

  • Bachelor's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 6+ years related experience (e.g., statistics, predictive analytics, research)
  • Master's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 4+ years related experience (e.g., statistics, predictive analytics, research)
  • Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 3+ years related experience (e.g., statistics, predictive analytics, research)
  • equivalent experience

Nice to have

  • Master's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 9+ years related experience (e.g., statistics, predictive analytics, research)
  • Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 6+ years related experience (e.g., statistics, predictive analytics, research)
  • equivalent experience
  • 5+ years of experience with developing and deploying machine learning solutions in production, with proven ownership of complex projects end-to-end (from problem formulation and data acquisition to model deployment and monitoring).
  • 3+ years of technical leadership experience in an applied science or data science team setting – this could include leading a team of scientists or acting as the key technical decision-maker on cross-discipline projects, with responsibility for delivering major features or systems.
  • Extensive hands-on expertise in ML techniques for predictive analytics, pattern recognition, and optimization. You should be comfortable selecting and tuning algorithms for regression, classification, clustering, time-series forecasting, etc., and understand their trade-offs

What the JD emphasized

  • define scientific vision
  • lead the development
  • architect end-to-end machine learning pipelines
  • lead the incorporation of LLM-based components
  • guiding the development of prompt frameworks, fine-tuning strategies, and retrieval-augmented techniques
  • ensure robust evaluation and governance of all AI/ML solutions
  • track emerging trends in AI

Other signals

  • integrates advanced machine learning models with emerging agentic capabilities powered by large language models (LLMs)
  • model recommendations, automate analysis, generate contextual summaries, and streamline workflows
  • define scientific vision and lead the development of both ML and LLM components
  • translate research into reliable production systems
  • define and drive the modeling strategy for the advertising recommendations platform, spanning classical machine learning (for analytics on structured data) and the use of generative AI
  • architect end-to-end machine learning pipelines
  • lead the incorporation of LLM-based components for the platform’s intelligent narrative generation
  • guiding the development of prompt frameworks, fine-tuning strategies, and retrieval-augmented techniques
  • ensure robust evaluation and governance of all AI/ML solutions
  • track emerging trends in AI