Principal Applied Scientist, Foundation Models, Agents & Decision Intelligence

About the role

Microsoft AI Research team focuses on building trustworthy AI systems that understand entities, content, and behavior over time. The Principal Applied Scientist leads scientific initiatives in foundation models, agents, and decision intelligence, mentoring scientists and shaping architecture for AI-driven trust and safety. Bengaluru, Karnataka, India, on-site.

What you’ll do

• Define and lead scientific initiatives in foundation models, agents, and decision intelligence
• Develop scalable learning systems for uncertainty propagation and threat modeling
• Mentor scientists and influence AI architecture across product and security teams
• Translate research insights into data strategies, model architectures, and evaluation plans
• Advance training, post‑training, and evaluation of tool‑using agents

What you’ll bring

• PhD or Master's in Computer Science or related quantitative field
• 10+ years of ML research experience
• Expertise in probability, statistics, and optimization
• Proficiency in Python and ML frameworks (PyTorch, JAX, TensorFlow)
• Experience with foundation/representation learning and uncertainty modeling
• Track record of deploying large-scale models in production
• Leadership in scientific direction and mentorship
• Publications in machine learning, agents, or trust and safety

Skills