Top AI & Machine Learning Jobs in San Francisco, CA

Enable frontier AI models on Microsoft's Maia AI accelerators by integrating model architectures, optimizing performance, improving serving capabilities, and building validation and automation infrastructure. Diagnose issues across software, systems, and hardware layers; support hardware bring-up; and partner with model, runtime, infrastructure, and hardware teams to deliver reliable, scalable production AI workloads.

Define technical strategy for enabling AI models and hardware generations on Microsoft's Maia accelerators. Lead cross-stack investigations across models, serving systems, infrastructure, and hardware; improve performance, scalability, reliability, validation, and automation; and align teams on durable platform capabilities and long-term technical investments.

Leads a team of cloud and AI solution architects delivering Copilot, agent, and cloud-native business solutions. Responsibilities include talent development, customer and partner engagement, solution governance, risk management, technical standards, and execution of AI transformation initiatives. The role drives customer adoption, time-to-value, and ROI while shaping innovative solutions for professional services organizations.

Develop and optimize AI compilers and code generation for new hardware, using LLVM/MLIR, Torch Dynamo, and Triton. Implement compiler features and optimizations in C++ and Python, improve AI workload performance, design programming abstractions, collaborate with hardware and ML teams, and plan and deliver engineering solutions.

Lead architecture and hands-on implementation for AI frameworks, performance systems, benchmarking, compilers, runtimes, and developer tools. Drive cross-team initiatives that improve LLM onboarding, inference performance, reliability, hardware utilization, and Azure capacity efficiency. Conduct complex systems investigations, establish scalable measurement and observability, influence technical strategy, mentor engineers, and deliver reusable production platforms across models, infrastructure, and silicon.

Design, implement, test, and operate production AI framework, runtime, benchmarking, performance, automation, observability, and developer-tooling components. Benchmark and optimize large language model training and inference across GPUs and Microsoft hardware. Diagnose cross-stack performance and reliability issues, drive projects through production deployment, collaborate with research, infrastructure, model, and hardware teams, and mentor engineers.

Develop and lead production AI infrastructure across frameworks, runtimes, benchmarking, performance tooling, observability, and hardware platforms. Responsibilities include optimizing large language model training and inference, building automation, diagnosing cross-stack performance issues, defining technical architecture, driving platform improvements, collaborating across research and infrastructure teams, and mentoring engineers. The role spans senior and principal-level individual contributor responsibilities focused on scalable, reliable, and efficient AI systems.

Lead end-to-end data science projects in a forward-deployed model: acquire and prepare messy datasets, develop and evaluate production-ready models tied to business outcomes, run experiments, drive ethics/privacy discussions, define solution strategy, and coach cross-disciplinary teams while applying latest AI approaches.

Lead end-to-end data science delivery embedded with customers: acquire and prepare messy data, develop and evaluate production-grade models tied to business outcomes, run experiments, drive ethics/privacy discussions, coach cross-functional teams, and define solution and customer strategy while applying cutting-edge AI and scalable operational modeling.

Build and operate large-scale GPU infrastructure for AI training and inference. Develop workload scheduling, serving, orchestration, failover, resource governance, and Kubernetes-based services using C# and Go. Optimize reliability, performance, utilization, observability, and diagnostics across global GPU fleets. Provide technical leadership, influence architecture, and collaborate across infrastructure, hardware, networking, and AI platform teams.