Staff Engineer, ML Profiling Tools

Description

The AGI Computing Lab works on the system-level problems created by growing AI/ML workloads.

The team designs scalable platforms that meet the compute and memory demands of these workloads while using less energy and delivering more performance.

Engineers in the lab work closely with hardware and software colleagues, test new computing abstractions, and research memory, computing, interconnect, and AI/ML trends.

The goal is to change how AI/ML applications are deployed and run, supporting progress toward AGI in an affordable and sustainable way.

The company makes technology used in smartphones, electric vehicles, hyperscale data centers, and IoT devices.

Responsibilities

• Build and maintain large, complex, highly interactive web-based visual analytics tools for AI/ML workloads.
• Create visualizations such as flame graphs, call trees, and timeline charts that display massive, multi-dimensional infrastructure telemetry smoothly.
• Work with compiler engineers and hardware architects to turn dense performance profiling metrics into clear developer workflows.
• Study and improve the ML performance profiling experience, designing tools and workflows that help engineers find and remove inference bottlenecks.
• Communicate with users, partners, and management so systems ship on schedule and within budget.
• Take on other duties as assigned.

Requirements

• A bachelor's degree with 10+ years, a master's degree with 8+ years, or a PhD with 5+ years of industry experience.
• Extensive experience building rich, interactive data visualizations and user interfaces with modern web technologies.
• Background in, or interest in, developer tools, IDE extensions, or complex diagnostic dashboards.
• Proven ability to turn large, unstructured, or multidimensional telemetry and profiling data into readable visuals.
• A track record of designing workflows for technical users such as software engineers, data scientists, or researchers, reducing cognitive load and speeding root-cause diagnosis.
• Hands-on experience with, or strong curiosity about, profiling and tracing tools such as Perfetto, Chrome Tracing, NVIDIA Nsight Systems/Compute, PyTorch Profiler, TensorBoard, or eBPF.
• Basic understanding of CPU, GPU, TPU, and interconnect/networking architectures and common bottlenecks such as memory bandwidth, compute utilization, and synchronization stalls.
• Bonus: Working knowledge of ML frameworks such as PyTorch, JAX, or TensorFlow, and execution paradigms such as LLM inference, pipeline/tensor parallelism, or kernel dispatch.
• Strong problem-solving skills and the ability to think critically and creatively.
• Inclusive, adjusting your style to each situation and to diverse global norms.
• A curious, resilient learner who looks for data to build understanding.
• Collaborative, builds relationships, offers support humbly, and welcomes other approaches.
• Innovative, explores new ideas, and adapts quickly to change.

Conditions

Daily onsite presence at the San Jose, CA office, the U.S. headquarters, under the Flexible Work policy.

Base pay range of $163,000 - $253,000 USD for this level across all US locations and functions.

Pay within the range varies by work location and may depend on job-related knowledge, skills, and experience.

Benefits

• Incentive opportunities that reward individual and company performance.
• Medical, dental, vision, and 401k, plus an inclusive rewards plan supporting physical and personal wellbeing.
• A charitable giving match and regular chances to get involved in the community.
• 4+ weeks of paid time off per year, plus holidays and sick leave.
• A stipend for fertility care or adoption, caregiver support, and medical travel support.
• On-demand apps and free confidential therapy sessions.
• An onsite café and gym, plus virtual classes.
• A flexible environment to help you find the right balance.