Senior Researcher - CoreAI
Job Description
The Microsoft CoreAI Post-Training team is dedicated to advancing post-training methods for both OpenAI and open-source models. Their work encompasses continual pre-training, large-scale deep reinforcement learning running on extensive GPU resources, and significant efforts to curate and synthesize training data. In addition, the team employs various fine-tuning approaches to support both research and product development. The team also develops advanced AI technologies that integrate language and multi-modality for a range of Microsoft products. The team is particularly active in developing code-specific models, including those used in Github Copilot and Visual Studio Code, such as code completion model and the software engineering (SWE) agent models. The team has also produced publications as by-products, including work such as LoRA, DeBerTa, Oscar, Rho-1, Florence, and the open-source Phi models. We are looking for a Senior Researcher - CoreAI with significant experience in large-scale model training, data curation, and hands-on coding, ideally from leading research labs. You will develop LLMs, SLMs, multimodal models, diffusion models, agentic models, and coding models using both proprietary and open-source frameworks. Key responsibilities include improving model quality and training efficiency through advanced techniques and data strategies, and managing the full pipeline from data ingestion, evaluation, to inference. Our team values startup-style efficiency and practical problem-solving. We are seeking a curious, adaptable problem-solver who thrives on continuous learning, embraces changing priorities, and is motivated by creating meaningful impact. Candidates must be self-driven, able to write efficient code and debug training jobs, document findings, and demonstrate a track record in these fields. You may include information about any individuals who can serve as your referral in your application. 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 positively impact our culture every day.
Locations
- Redmond, Washington, United States, Redmond, Washington, United States
- Sunnyvale, California, United States, Sunnyvale, California, United States
Salary
Salary not disclosed
Required Qualifications
- Doctorate in relevant field OR equivalent experience. (degree)
- OR equivalent experience. (degree)
- Publication record with over 1000 citations (degree)
- 2+ years of experience in large-scale model training, especially with LLMs, SLMs, multimodal, or code-specific models (degree)
- 2+ years of expertise in data curation and synthesis, creating and refining datasets to optimize training outcomes (degree)
- 2+ years of coding experience in languages such as Python as well as frameworks suhc as PyTorch and Triton with the ability to write efficient, research or production code and debug complex training jobs (degree)
- 2+ years of experience with both proprietary and open-source frameworks with demonstrated proficiency in training pipelines and architecture (degree)
- Microsoft Cloud Background Check: This position will be required to pass the Microsoft Cloud background check upon hire/transfer and every two years thereafter. (degree)
- Proven track record of impactful research, preferably at leading research labs, with published work or real-world deployments (degree)
- Extensive experience with foundation models, including large-scale training, model inference, reinforcement learning, reasoning models, vision-language integration, and audio-visual modeling (degree)
- Hands-on experience with large-scale distributed training or serving, and systems of thinking (degree)
- Proficiency in programming languages such as Python, and experience with machine learning frameworks like PyTorch and Triton (degree)
- Experience working with large, complex datasets and developing data pipelines for LLM training (degree)
- Demonstrated ability to collaborate within interdisciplinary teams and communicate complex, multimodal research concepts effectively (degree)
- Startup-style mindset, be agile, solution-oriented, and able to operate with minimal overhead (degree)
- Self-driven and organized with the ability to take ownership of projects and document findings clearly and effectively (degree)
Responsibilities
- Perform large-scale model training — Especially with LLMs, SLMs, multimodal, or code-specific models.
- Perform data curation and synthesis — Creating and refining datasets to optimize training outcomes.
- Hands-on coding— Write efficient, production-quality code and debug complex training jobs.
- Work on both proprietary and open-source frameworks — Demonstrated proficiency in training pipelines and architecture.
- Full-stack modeling responsibility — From data ingestion and training to evaluation and inference management.
- Contribute to or build on existing innovations like technical report of the well-known models.
- Develop novel AI solutions that bridge language, vision, and code understanding.
- Help develop models powering tools like GitHub Copilot, Cursor, and VS Code suggestions.
- Embody our Culture and Values
Travel Requirements
Microsoft on-site only
