Microsoft
Senior Applied Scientist
$120K–$261K
Redmond, WA · On-site · Senior · Full-time · Posted Oct 14, 2026
- Experience
- Required: 4+ years with a bachelor’s degree, 3+ years with a master’s degree, or 1+ year with a doctorate. Preferred: 6+ years with a master’s degree, 3+ years with a doctorate, or equivalent experience.
- Education
- Bachelor’s, master’s, or doctorate in Statistics, Computer Science, Electrical or Computer Engineering, or a related field. Preferred qualifications also state “OR equivalent experience.”
About the job
Develops scientific methods and production capabilities to make Copilot offline evaluations more representative of customer experiences and predictive of online results.
What you'll do
- Develop and evaluate machine learning methods, synthetic evaluation data, and experimentation frameworks for Copilot.
- Run offline and online experiments, analyze results, and recommend improvements to evaluation quality and alignment.
- Partner with engineering and product teams to deliver scalable production capabilities; provide technical leadership and mentoring.
What they ask for
- Bachelor’s degree in a listed or related field plus 4+ years of related experience, or master’s plus 3+ years, or doctorate plus 1+ year.
- Meet Microsoft, customer and/or government security screening requirements, including the Microsoft Cloud Background Check.
Nice to have
- Production machine learning systems experience, or experience in relevant areas such as generative AI, LLMs, NLP, or responsible AI.
- Experience in Privacy-Preserving ML, Synthetic Data Generation, and LLM/Agentic Evaluation.
Pay
USD $119,800 - $234,700 per year (base salary); USD $160,200 - $261,000 per year (base salary)
Tools
machine learning, AI, generative AI, LLM, natural language processing, recommender systems, computer vision, synthetic data
AI focus
Applies machine learning and generative AI to Copilot evaluation, including representative and synthetic evaluation data, agent and LLM evaluation, and improving agreement between offline evaluations and online shadow experiments.
Model-written summary · the original posting is authoritative