Principal Data Science Manager

Microsoft Microsoft · Big Tech · Herzliya, Tel Aviv District, IL +1 · Data Science

Lead a team developing production-oriented AI solutions for computer vision and small language models, focusing on efficiency for edge devices on Windows products. This role involves managing researchers, MLOps engineers, and data management experts, defining technical roadmaps, and driving AI solutions from research to production deployment.

What you'd actually do

  1. Lead, grow, and develop a diverse data science team that includes researchers, MLOps engineers, and data management experts, while fostering a culture of technical excellence, collaboration, ownership, and continuous learning.
  2. Define and drive the team’s technical roadmap in close partnership with product and system engineering teams.
  3. Guide the development of production-oriented AI solutions from research exploration through production deployment, with a strong focus on quality, scalability, reliability, and measurable customer impact.
  4. Drive model efficiency and deployment readiness for constrained environments, ensuring AI solutions meet real-world requirements for performance, latency, footprint, and resource usage.
  5. Mentor team members, create clarity, set priorities, and help the team deliver high-impact results in a dynamic and evolving environment.

Skills

Required

  • M.Sc. or Ph.D. in Computer Science, Electrical Engineering, or a related technical field.
  • Experience in computer vision, efficient small language models, or related AI model development areas
  • Experience as a people manager or as a technical leader driving work across a team.

Nice to have

  • Hands-on experience with computer vision systems deployed in production or resource-constrained environments.
  • Hands-on experience with model development, evaluation, optimization, compression and efficient inference for edge devices.
  • Experience bringing AI solutions from research or prototyping into production environments.
  • Experience working with MLOps, data pipelines, model lifecycle management, or production AI infrastructure.
  • Demonstrated ability to motivate people, create clarity, and drive execution.
  • High motivation, ownership mindset, strong collaboration skills, and the ability to influence across disciplines.

What the JD emphasized

  • production-oriented AI solutions
  • real-world edge-device constraints
  • efficiently under real-world edge-device constraints
  • model efficiency and deployment readiness for constrained environments
  • AI solutions meet real-world requirements for performance, latency, footprint, and resource usage.

Other signals

  • production-oriented AI solutions
  • advanced models to Windows products
  • state-of-the-art computer vision solutions
  • efficiently under real-world edge-device constraints
  • SLMs, enabling advanced AI models to run more efficiently