Perception Engineer, Autonomy (Air Domain)

Description

The company builds the digital infrastructure that lets intelligence run on any moving machine, a field it calls physical AI. Founded in 2017 and valued at $15 billion, it is based in Sunnyvale, California.

It serves the automotive, defense, trucking, construction, mining and agriculture industries through three areas: tools and infrastructure, operating systems, and autonomy. 18 of the top 20 global automakers, plus the United States military and its allies, rely on its solutions.

Other offices are in Washington, D.C.; San Diego; Ft. Walton Beach, Florida; Ann Arbor, Michigan; London; Stuttgart; Munich; Stockholm; Bangalore; Seoul; and Tokyo.

The role is on the perception team for autonomy on uncrewed air vehicles. The team turns noisy, intermittent data from many sensors and teammates into a dependable view of the airspace.

The work covers classical state estimation and tracking, sensor fusion, and learning-based prediction. These models increasingly adapt during an engagement, using limited and delayed feedback. You will write real-time C++ algorithms, build tooling and simulation studies to test them, and move research prototypes into production software. You need fluency in both estimation theory and machine learning, rigor in measuring results, and comfort moving between research and production.

Responsibilities

• Develop, integrate and adapt AI/ML algorithms on the perception autonomy stack that gather useful information from the platform's sensor suite.
• Write software that controls sensors such as EO, IR, acoustics, radar and RF automatically.
• Work with the sensor, tracking and autonomy teams so heterogeneous platform swarms deploy smoothly at on-site DoD testing and demonstration events.
• Ship containerized software to embedded Linux devices, using computer-in-the-loop testing and profiling, and collect performance data efficiently.
• Speak with the DoD customer during field events to learn their use cases, requirements and triage needs, and help deliver a strong customer experience.

Requirements

• BS or MS in Electrical, Aerospace or Mechanical Engineering, Computer Science, Robotics or a related field, plus 3+ years building perception or estimation software for autonomous systems.
• A bachelor's degree in Computer Vision, Computer Science, Electrical Engineering, Robotics or a related technical field, plus 2+ years of industry experience.
• Strong modern C++ (17 or 20) for real-time, performance-critical code on Linux, with comfort in linear algebra libraries such as Eigen and in profiling.
• Solid grounding in state estimation, including Kalman filter variants (EKF and UKF), multiple-model filters (IMM), and covariance and uncertainty handling.
• Experience with multi-target tracking and sensor fusion: data association, track management, and fusing noisy heterogeneous sensors that report intermittently.
• Working knowledge of machine learning for prediction, including sequence models (RNN, LSTM, transformers), training pipelines, and evaluation on held-out data.
• Python for data pipelines and for analyzing and visualizing logged and simulated runs.
• Experience validating algorithms in simulation and on recorded data: designing experiments, defining metrics, and reporting results honestly.
• Experience with online, real-time learning, where models adapt during operation rather than only in offline training.
• Background in adaptive control or learning-based estimation, such as concurrent learning, Lyapunov-based learning, or Bayesian or recursive least-squares parameter estimation.
• Experience learning from intermittent, delayed or sparse feedback, where ground truth arrives late or only sometimes.
• Behavior and intent prediction for other agents, including models of non-cooperative or adversarial actors.
• Experience deploying small neural networks in C++ without heavy ML frameworks, under tight latency and compute budgets.

Conditions

Full-time employees are expected to work from the office 5 days a week.

Some flexibility is allowed: occasional remote work, starting the day with morning meetings from home before going in, or leaving early for family commitments.

The in-office expectation does not apply to contractor positions.

You must be a U.S. citizen and must hold or be eligible to obtain and maintain a U.S. security clearance.

Candidates who do not match every qualification are encouraged to apply for this or other roles.

US-based applicants may request accommodations during the application and interview process, which will not affect their candidacy, and they do not need to disclose their condition.

Benefits

• Eligible roles include base salary, equity, and benefits. Total pay may include equity as options and/or restricted stock units, comprehensive health, dental, vision, life and disability insurance, a 401k with employer match, learning and wellness stipends, and paid time off.
• Base salary is one part of total pay. Benefits may change and can vary by jurisdiction of employment.
• The posted pay range is the intended target base salary for new hires. The actual offer depends on experience, credentials, education, skill level, interview performance, and the level and scope of the position.