Researcher - Multiple Levels, Microsoft Research AI Frontiers

The Microsoft Research AI Frontiers Lab is on a mission to push the boundaries of what's possible in AI — advancing capabilities, efficiency, and safety through breakthroughs in agentic model and systems. Our previous work spanned Small Language Models (Phi, Orca), Synthetic Data Generation (AgentInstruct), Computer Use Models (Fara), and agentic AI systems (AutoGen, MagenticOne, OmniParser), with real-world impact at scale. 

We're looking for Senior and Principal Researchers to join us in Redmond, WA or New York City, NY and help shape the future of agentic AI. 

In this role, you'll develop and implement cutting-edge training methodologies for large language and multimodal models with advanced agentic and reasoning capabilities. Your work contribute to cutting edge research and directly influence state-of-the-art systems used by millions of people worldwide. 

Our current focus areas include: 

• Reasoning and inference-time compute
• Action models for automating computer use tasks 
• Memory and long-horizon task execution
• Multi-agent training and orchestration 

The AI Frontiers Lab is a uniquely dynamic environment: multidisciplinary by design, connected to top academic institutions, and committed to open publication. You'll have access to rich, diverse real-world data and the runway to experiment, publish, and make your mark on one of the most consequential fields in technology.

 At Microsoft, our mission—to empower every person and every organization on the planet to achieve more—guides how we partner with customers to deliver trusted, impactful solutions. With a growth mindset culture, we innovate responsibly and measure success by shared progress—people, teams, and customers. Join us to do meaningful work that changes the world and helps shape what’s next for everyone.

• Lead and conduct cutting-edge research in agentic AI, translating insights into practical solutions that ship to millions of users — in close collaboration with researchers, engineers, and product teams.
• Curate high-quality datasets and develop novel training algorithms tailored to agentic and reasoning tasks.
• Design rigorous evaluation frameworks to benchmark model capabilities across reasoning, long-horizon tasks, and multi-agent scenarios. 
• Prototype and iterate on new agentic capabilities — from computer use and action models to memory and multi-agent orchestration. 
• Drive research from idea to publication, with opportunities to publish at top-tier venues and contribute to the broader scientific community. 

Required Qualifications 

• Doctorate in Computer Science or relevant field 
  • OR Master's Degree in Computer Science or relevant field AND 3+ years related research experience
  • OR Bachelor's Degree in Computer Science or relevant field AND 4+ years related research experience
  • OR equivalent experience.

Preferred Qualifications 

• Doctorate of Computer Science or relevant field AND 2+ years related research experience 
  • OR equivalent experience. 
• Research program demonstrated by public artifacts like models, tools, code in AI space or publications at the following conferences: NeurIPS, ICML, ICLR, ACL, NAACL, CVPR, COLT, ECCV, ICCV, EMNLP.
• Experience with building and evaluating frontier models.
• Experience publishing academic papers as a lead author or essential contributor in the field of Artificial Intelligence.
• Experience participating in a top conference in relevant research domain.
• Demonstrable ability to define an ambitious, original research agenda.
• Ability to collaborate, communicate effectively, and technically lead multi-disciplinary team.
• Keen interest in real-world applications and impact. 

#Research