Microsoft at OSDI 2026
Lieu: Seattle, Washington
Systems research at Microsoft
Explore the breadth of systems and networking research at Microsoft—where world-class researchers and product teams collaborate to solve complex, real-world challenges at global scale. Discover how each team is advancing the state of the art.
Microsoft Research
Microsoft Research unites global interdisciplinary teams to advance systems research and foster impactful collaborations with academia and product teams for the benefit of customers, community, and society. By bringing together expertise in systems, networking, programming languages, machine learning, security, and privacy, Microsoft Research innovates on cloud, AI, and datacenter infrastructure. MSR’s focus is on developing solutions that are sustainable, reliable, efficient, and secure in the agentic AI era.
MSR Asia
- MSRA (Shanghai) System and Engineering Group
- Systems and Networking Research Group (Asia)
- Microsoft Research Asia – Vancouver
MSR Cambridge, UK
MSR India
CoreAI
CoreAI’s objective is to empower everyone to shape the future with AI. For developers, this means using AI-powered tools and agents to transform the entire software development lifecycle. For organizations, this means moving beyond first-generation AI applications and into a world of dynamic, agentic applications that are designed from the ground up to act on their behalf with intelligence, memory, adaptability, and security at their core. To drive this transformation, CoreAI is building a complete set of AI-powered developer tools, on top of a full-stack AI app and agent factory, with security and trust baked in from the start, that reaches from the cloud to the edge. Learn more on the GitHub blog (opens in new tab) and the Foundry blog (opens in new tab).
Azure Research
Azure Research conducts world-class research relevant to Azure. The Security and Privacy group focuses on cloud security and privacy, shaping Microsoft’s products and services through research on secure distributed computing, privacy-preserving AI, software security, and verified cryptography. Meanwhile, the Systems group enhances the cost efficiency of Microsoft’s online services and datacenters by exploring chip-to-datacenter power management, efficient AI infrastructure, sustainability, resource optimization, next-generation cloud platforms, and AI-driven systems, driving technological innovation and operational efficiency for Azure.
M365 Research
M365 Research is an applied research organization advancing enterprise-grade Agentic AI to evolve Microsoft Copilot into a trusted and skilled digital coworker embedded in the flow of enterprise work. The group’s technical roadmap is structured around two pillars: Agentic Innovation, which advances reasoning, planning, multi-agent collaboration, and persistent memory systems with robust self-correction and valuation mechanisms to reliably deliver high-quality enterprise work outcomes; and Efficient AI, which focuses on inference optimization, systems and hardware-software co-design, and latency and cost efficiency improvements to enable responsive, planet-scale deployment across the Microsoft 365 ecosystem. Together, these pillars ensure Agentic AI systems are both capable of producing trusted work outcomes and efficient enough to operate reliably across real-time enterprise environments at global scale.