We own inference performance of OpenAI and other state of the art LLM models and work directly with OpenAI on the models hosted on the Azure OpenAI service serving some of the largest workloads on the planet with trillions of inferences per day in major Microsoft products, including Office, Windows, Bing, SQL Server, and Dynamics. Lead technical alignment across teams by shaping architecture and priorities, resolving competing requirements, and holding researchers, product groups, infrastructure owners, and hardware partners accountable to clear decisions, execution plans, and measurable outcomes.