New 
Principal Researcher - CoreAI
 Microsoft | |
 United States, Washington, Redmond | |
 Jul 13, 2025 | |
|
OverviewThe Microsoft CoreAI Modeling team develops advanced AI technologies that integrate language and multi-modality for various Microsoft products. We built models for internal and external applications, focusing on post-training OpenAI models, OSS models through large-scale continual pre-training, and RL, as well as code-specific models such as SWE. We are seeking a Principal Researcher - CoreAI with expertise and experience in large-scale modeling training and data curation. You will work on LLMs, SLMs, multimodal systems, and coding models using proprietary and open-source frameworks. This will include managing the full model pipeline from dataset ingestion to training, evaluation, and inference. Our team operates like a startup, emphasizing efficiency and hands-on problem-solving. Applicants must write code, debug training jobs, and document their learning experiences. Microsoft's mission is to empower every person and every organization on the planet to achieve more. As employees we come together with a growth mindset, innovate to empower others, and collaborate to realize our shared goals. Each day we build on our values of respect, integrity, and accountability to create a culture of inclusion where everyone can thrive at work and beyond. In alignment with our Microsoft values, we are committed to cultivating an inclusive work environment for all employees to positivelyimpactour culture every day.
ResponsibilitiesDesign and evolve our distributed-training platform that supports LLMs, SLMs, multi-modal, code and open-source models. Write production-grade code that automates data preprocessing, sharding, checkpointing, mixed-precision & pipeline parallelism. Profile and monitoring end-to-end training runs, eliminate bottlenecks, tune kernels, and devise new scheduling and placement strategies. Data preparation, training, and evaluation of customization tasks. Collaboration with Microsoft product groups to integrate multimodal AI solutions across applications. Staying updated with the latest advancements in deep learning, inference optimization, and multimodal learning. Collaborate with top students and interns with opportunity to publish great work to community. Embody our Culture and Values | |