Senior Manager, AI Deployment

General Motors•Sunnyvale, CA

10d•$296,300 - $453,900•Remote

About The Position

General Motors is developing the software and artificial intelligence capabilities for the next generation of autonomous driving. Within AI Foundations, AI Acceleration makes machine learning models faster, more efficient, and more reliable on production vehicle hardware. The AI Deployment team owns model inference performance across simulation, hardware-in-the-loop, bench, and vehicle environments. The team focuses on latency, memory, GPU utilization, numerical parity, profiling, benchmarking, reduced precision, and production readiness. We are looking for a Senior Manager, AI Deployment to lead the strategy and execution of model performance and on-vehicle inference for autonomous driving. You will lead engineering managers and senior technical leaders working across model optimization, GPU systems, inference runtimes, and vehicle integration. You will set performance goals, guide optimization of complex autonomy models, and establish disciplined methods to measure latency, diagnose regressions, and validate improvements. Success requires strong technical judgment, people leadership, and the ability to make clear trade-offs among latency, memory, throughput, accuracy, power, and numerical parity.

Requirements

Bachelor’s degree in Computer Science, Electrical or Computer Engineering, Robotics, Machine Learning, or a related field; advanced degree preferred, or equivalent experience.
10+ years of experience in machine learning systems, model optimization, inference, GPU systems, robotics, autonomous driving, or a related field.
5+ years of people-leadership experience, including experience leading managers or senior technical leaders.
Experience shipping production machine-learning inference systems on GPU, accelerator, robotics, automotive, or other edge hardware.
Strong understanding of the factors that determine model performance: architecture, tensor shapes, operators, kernels, memory movement, scheduling, runtime execution, and hardware utilization.
Hands-on experience with several of the following: PyTorch, CUDA, C++, Python, TensorRT, GPU profiling, benchmarking, performance analysis, or inference runtimes.
Experience with quantization, pruning, distillation, architecture optimization, kernel optimization, or memory optimization.
Experience building benchmark automation, performance regression detection, telemetry, dashboards, or profiling workflows.
Strong systems thinking, communication, decision-making, and cross-functional leadership skills.

Nice To Haves

Experience optimizing real-time machine-learning systems for autonomous driving, robotics, embedded systems, or computer vision.
Deep experience with GPU performance, memory bandwidth, occupancy, synchronization, stream scheduling, or device-to-device data movement.
Experience with NVIDIA Nsight Systems, NVIDIA Nsight Compute, PyTorch Profiler, TensorRT profiling tools, or equivalent tools.
Experience deploying reduced-precision models and managing calibration, sensitivity, parity, and model-quality risks.
Experience optimizing transformer, vision, lidar, or multimodal workloads.
Experience measuring performance across simulation, hardware-in-the-loop, bench, and vehicle environments.
Experience with safety-critical or highly reliable systems.

Responsibilities

Own the strategy, roadmap, and operating plan for AI model performance and inference quality.
Establish performance budgets for latency, throughput, memory, GPU utilization, power, and numerical parity.
Lead investigations into performance bottlenecks across model architecture, operators, kernels, memory movement, scheduling, runtime behavior, and hardware utilization.
Establish repeatable benchmarking and profiling practices across simulation, hardware-in-the-loop, bench, and vehicle environments.
Guide optimization through model architecture changes, operator and kernel improvements, memory optimization, scheduling, and hardware-aware execution.
Build performance dashboards, regression detection, benchmark automation, and root-cause diagnostics.
Partner with Embodied AI, model development, GPU kernel, runtime, system performance, vehicle integration, simulation, and safety teams.
Influence model design by translating profiling results into clear recommendations for model architects and researchers.
Represent AI Deployment in architecture reviews, program planning, and senior leadership discussions.
Build and lead an inclusive, high-performing organization through hiring, coaching, feedback, and manager development.
Establish clear ownership, priorities, staffing plans, and operating rhythms across performance workstreams.
Define and manage KPIs for inference latency, latency variability, throughput, memory efficiency, GPU utilization, parity, and regression rate.
Balance near-term production needs with longer-term investments in profiling, optimization automation, reduced precision, and performance infrastructure.
Resolve cross-functional issues and align stakeholders when performance, quality, or implementation trade-offs are contested.
Develop technical leaders and succession plans in GPU performance, model optimization, inference systems, and numerical analysis.