Microsoft AI is seeking a Software Engineer for AI infrastructure as part of the MAI Superintelligence Team to help build the next wave of capabilities for Copilot and other consumer AI products. The successful candidate will contribute across engineering, research, and deployment activities, including model architecture, data curation, training and inference infrastructure, evaluation protocols, RLHF, and related areas. The role involves benchmarking and tuning training and inference on GPU clusters, instrumenting tooling, and collaborating with model scientists to ensure reliable, scalable AI jobs in production.

Responsibilities include: - Develop and tune the pretraining scalable software for Nvidia GB200 72NVL CX8 and AMD MIxxx architectures. - Benchmark GB200 and AMD MIxxx GPU clusters. - Gather data and insights to shape the pretraining compute roadmap. - Care deeply about conversational AI and its deployment. - Actively contribute to development of AI models powering our products. - Navigate obstacles to deliver work to users quickly and iteratively. - Thrive in a fast-paced, design-driven product development cycle. - Embody our culture and values.

Qualifications: - Bachelor's degree in Computer Science or related field, plus 4+ years of technical engineering experience with programming in C/C++/C#/Java/JavaScript/Python, or equivalent. - OR 8+ years of experience; or a Master's with 6+ years of experience. - Experience with generative AI and distributed computing. - Experience leading technical projects and making data-driven architectural decisions.

Compensation

Microsoft AI is seeking a Software Engineer for AI infrastructure as part of the MAI Superintelligence Team to help build the next wave of capabilities for Copilot and other consumer AI products. The successful candidate will contribute across engineering, research, and deployment activities, including model architecture, data curation, training and inference infrastructure, evaluation protocols, RLHF, and related areas. The role involves benchmarking and tuning training and inference on GPU clusters, instrumenting tooling, and collaborating with model scientists to ensure reliable, scalable AI jobs in production.

Responsibilities include: - Develop and tune the pretraining scalable software for Nvidia GB200 72NVL CX8 and AMD MIxxx architectures. - Benchmark GB200 and AMD MIxxx GPU clusters. - Gather data and insights to shape the pretraining compute roadmap. - Care deeply about conversational AI and its deployment. - Actively contribute to development of AI models powering our products. - Navigate obstacles to deliver work to users quickly and iteratively. - Thrive in a fast-paced, design-driven product development cycle. - Embody our culture and values.

Qualifications: - Bachelor's degree in Computer Science or related field, plus 4+ years of technical engineering experience with programming in C/C++/C#/Java/JavaScript/Python, or equivalent. - OR 8+ years of experience; or a Master's with 6+ years of experience. - Experience with generative AI and distributed computing. - Experience leading technical projects and making data-driven architectural decisions.

Compensation