Senior Systems Engineer, Event Detection and Response

About the role

The Senior Systems Engineer, Event Detection and Response owns the full event lifecycle for autonomous operations, serving as the critical link between raw data and fleet safety decisions. You will shape intake rules, construct detectors, review signals, deploy insights, and collaborate closely with cross-functional peers to ensure timely and accurate responses. This position translates complex log data into decisive action while maintaining rigorous traceability and prioritizing public safety above all else. You will define the boundaries of what matters, build the tools that see it, and own the narrative from detection to resolution. The role demands intellectual rigor, a bias for action, and comfort with ambiguity in a rapidly evolving autonomy stack. You will act as the technical owner for event-driven quality, ensuring that anomalies are not just seen but understood and resolved. This is a hands-on leadership position where your work directly determines the reliability and trustworthiness of our autonomous vehicles in diverse urban environments.

Key facts

What you'll do

Design and implement detection logic for rare and high-risk vehicle behaviors using both high-volume streaming data and deep historical logs.

Construct intake filters that isolate critical events while aggressively suppressing noise across varied operating cities and operational conditions.

Build robust detectors that convert raw logs from complex sources into structured, actionable signals for safety and data teams.

Collaborate with data engineering and simulation teams to embed detectors into CI pipelines and enrich scenario libraries for continuous validation.

Review surfaced events in simulation and regression suites to verify fixes before exposing changes to the fleet, ensuring safety is never compromised.

Ship detector updates that shorten the cycle time from initial detection to verified correction, enabling rapid learning and improvement.

Analyze large-scale fleet datasets to reveal patterns, frequency distributions, and risk profiles that inform strategic decisions at the leadership level.

Define system requirements and ensure end-to-end traceability from stakeholder needs through detector logic and into test scenarios and regression suites.

Analyze structured logs such as ROS bags, Parquet, Arrow, or proprietary formats to surface anomalies and build reliable signal pipelines.

Recreate log-derived events as parameterized scenarios in simulation platforms like CARLA or internal tools to validate detector behavior and edge cases.

Leverage Python (pandas, numpy, polars) and SQL to analyze large datasets, while demonstrating comfort with distributed platforms and data-intensive workflows.

Operate confidently in Linux environments, using command-line tools and source control systems such as git to manage complex detector configurations.

Translate detected anomalies into test cases that drive regression coverage and contribute to safety cases, strengthening the overall validation strategy.

Partner with cross-functional stakeholders to prioritize event types, align on risk tolerance, and maintain clear communication about detector performance.

Continuously refine detection rules and thresholds based on fleet feedback, ensuring that the system adapts to new patterns and operating environments.

Requirements

Candidates must bring 3 to 5 years of experience in automotive or robotics engineering, with a strong foundation in system-level design and operations.

At least 2 of those years should be dedicated to autonomous driving systems, with direct exposure to sensor suites, perception pipelines, and prediction and planning behavior.

You must understand how perception and planning outputs appear in logs and how to correlate them to vehicle state and environment events.

The ability to craft signal or rule-based detectors against structured logs is essential, using formats such as ROS bags, Parquet, Arrow, or proprietary systems.

Experience using simulation platforms like CARLA or internal tools to recreate log-derived events as parameterized scenarios is required.

You must analyze large datasets with Python (pandas, numpy, polars) and SQL, demonstrating fluency in data manipulation and exploratory analysis.

Comfort with distributed platforms and data pipelines is beneficial, as is experience operating within Linux-based production environments.

You will operate confidently in a command-line driven workflow, using source control such as git to manage detector definitions, tests, and versioned configurations.

Nice to have

Experience translating detected anomalies into test cases that drive regression coverage and safety cases is valued, though no specific tools are listed beyond those required in the qualifications.

Practical notes

This is a full-time engagement based in Ann Arbor, MI.

Compensation for this role ranges from $123,700.00 to $161,000.00 annually.

No specific visa or travel details are provided in the source material.

The listed work hours are not explicitly stated but align with standard full-time expectations for this role.