Senior Systems Performance Engineer

Motional•Pittsburgh, PA

2d•$149,000 - $198,500•Hybrid

About The Position

Motional’s Platform Infrastructure organization builds and operates the foundational onboard software systems that power our autonomous vehicles. Our responsibilities include system communications and IPC, networking, Linux operating system development and maintenance, high-speed data logging and offload, and onboard system performance optimization. We are seeking a Linux Systems Performance Engineer to improve the performance, reliability, and observability of our onboard compute platform. This will be a hybrid role focused on both the development of new observability methods and metrics while also operating with the rest of the engineering organization to diagnose and develop solutions to improve our real-time system. You will build tools and telemetry that enable faster, and increasingly real-time, root cause analysis while partnering with autonomy teams to diagnose and resolve resource contention, crashes, latency spikes, and other complex system-level issues.

Requirements

Strong experience debugging and optimizing Linux system performance.
Deep understanding of operating system concepts, including process scheduling, memory management, filesystems, networking, and inter-process communication.
Experience diagnosing resource contention, latency, crashes, or throughput problems in complex production systems.
Experience with Linux performance and tracing tools such as `perf`, ftrace, eBPF/BPF, BCC, bpftrace, flame graphs, or equivalent technologies.
Proficiency in a systems programming language such as C or C++.
Experience using Python, shell scripting, or similar languages for automation and data analysis.
Ability to analyze large volumes of system telemetry and distinguish root causes from symptoms.
Ability to work effectively across organizational boundaries and collaborate with application, infrastructure, and hardware teams.
Strong written and verbal communication skills, including the ability to explain complex system behavior to varied audiences.

Nice To Haves

Experience optimizing embedded Linux or automotive compute platforms.
Experience with real-time or latency-sensitive systems.
Experience working directly with the Linux kernel, scheduler, device drivers, or kernel instrumentation.
Experience designing observability platforms, telemetry pipelines, or fleet-scale performance analysis systems.
Experience with cgroups, namespaces, containers, and resource isolation.
Experience analyzing performance across heterogeneous compute resources such as CPUs, GPUs, and hardware accelerators.
Experience with high-throughput networking, storage, data logging, or data offload systems.
Experience establishing performance budgets, service-level objectives, or automated regression detection.
Experience with the NVIDIA DRIVE AGX platform or DriveOS ecosystem.
Familiarity with Bazel and large-scale C++ codebases.

Responsibilities

Design and build low-overhead observability infrastructure for resource usage, latency, scheduling behavior, and system health.
Develop metrics, tracing, profiling, and analysis tools that identify performance contention and explain system behavior.
Improve real-time and post-event root cause analysis across CPU, memory, storage, networking, and other shared resources.
Partner with autonomy and platform teams to investigate crashes, latency spikes, missed deadlines, resource exhaustion, and performance regressions.
Analyze application, operating system, kernel, and hardware interactions on embedded Linux compute platforms.
Translate investigations into durable improvements, including tooling, alerts, dashboards, performance requirements, and architectural changes.
Define performance benchmarks and regression tests for onboard systems.
Evaluate the performance impact of new software, hardware, and system configurations.
Communicate findings clearly and provide actionable recommendations to cross-functional engineering teams.
Contribute to technical roadmaps for onboard observability and system performance.