Principal Research Engineer – Generative AI – AI Frontiers - Microsoft : Job Details

Redmond, Washington, United States

Share job

Date posted

Jul 25, 2025

Job number

1852398

Work site

Up to 50% work from home

Travel

None

Role type

Profession

Research, Applied, & Data Sciences

Applied Sciences

Employment type

Full-Time

The AI Frontiers lab at Microsoft Research is charted with ambitious research goals for advancing Artificial Intelligence (AI) capabilities in several key areas including modeling, algorithms, reasoning, and agentic AI. Our lab offers a vibrant environment for cutting-edge multidisciplinary research, including an open publication policy and close links to top academic institutions around the world.

This Principal Research Engineer position offers a unique opportunity to contribute to some of the most challenging and rewarding problems in AI. You will develop novel ideas in reinforcement learning, and help evolve our codebases for pre-training, mid-training, and post-training, which have led to models like Phi, setting many new records. Collaboration across disciplines will be key to advancing reasoning and agentic AI.

Our lab provides opportunities for experimentation, access to diverse real-world problems and data, and the potential to deploy research to over a billion Microsoft customers. We aim to expand the capabilities, efficiency, and safety of AI through innovations in foundation models and learning agent platforms. Our projects include Phi, Orca, AutoGen, Eureka, OmniParser, Magnetic-One, Magnetic-UI, Dion, Belief State Transformers, among others.

Research areas include but are not limited to:

• Pre-training: language models, action models, multimodal models
• Mid-training: long context extension
• Post-training: instruction tuning, reinforcement learning from feedback
• Reasoning: scaling inference time via reinforcement learning
• Action Models: training models for digital actions (e.g., web agents)
• Orchestration and multi-agent systems: collaboration, human feedback, oversight

If your skills align with these areas, apply today and join us in this exciting journey!

Microsoft's mission is to empower every person and organization to achieve more. We foster a growth mindset, innovation, collaboration, and a culture of inclusion where everyone can thrive.

Required Qualifications

• Bachelor's in Computer Science or relevant field AND 6+ years related experienceOR Master's Degree in Computer Science or related field AND 4+ years related experienceOR Doctorate in Computer Science or related field AND 3+ years related experienceOR equivalent experience.
• At least 1 year of experience developing with Python and Pytorch/JAX.

Preferred Qualifications

• Familiarity with large language model architectures and optimizations
• Experience debugging and profiling distributed Pytorch code
• Understanding of CUDA kernels
• Experience with pre-training, mid-training, or post-training pipelines for language or multimodal models
• Foundational knowledge of reinforcement learning and its challenges
• Experience with verl, Ray, Megatron, or vLLM is a plus
• Building scalable services experience is beneficial

The typical base pay range in the U.S. is USD $139,900 - $274,800 per year, with higher ranges in specific locations like San Francisco and NYC. Benefits vary by role and location. Details at Microsoft Careers. Applications are accepted ongoing.

• Design, develop, and implement research projects in collaboration with teams.
• Contribute to breakthroughs in LLMs, reasoning, and agentic AI, helping improve and explore capabilities.

Benefits may vary by employment type and location, including healthcare, educational resources, discounts, investments, parental leave, time off, giving programs, networking opportunities, and more.

Microsoft is an equal opportunity employer. All qualified applicants will receive consideration without discrimination. For accommodations during the application process, read more here.