Senior Systems & Safety Engineer – AI/ML

Gatik AI • Santa Clara, CA

2d • $160,000 - $220,000 • Onsite

About The Position

Gatik is revolutionizing middle-mile logistics with fully driverless commercial box trucks and tractors operating on fixed B2B routes. As a Senior Systems & Safety Engineer for AI/ML, you will bridge the gap between cutting-edge deep learning algorithms (perception, prediction, planning) and rigorous safety engineering. You will define the functional safety and SOTIF strategies for learned components, establish out-of-distribution (OOD) detection criteria, design runtime AI safety monitors, and construct evidence-backed safety cases for machine learning models running on Level 4 autonomous commercial vehicles.

Requirements

• M.S. or Ph.D. in Computer Science, Robotics, Electrical Engineering, Machine Learning, or related field.
• 6+ years in systems/safety engineering or autonomous systems, with at least 3 years explicitly focused on AI/ML safety.
• Deep working knowledge of ISO 21448 (SOTIF), ISO/PAS 8800, UL 4600, and ISO 26262
• Strong understanding of sensor modalities (LiDAR, Radar, Cameras, IMUs) and their failure modes in adverse weather/environmental conditions.

Nice To Haves

• Track record of taking a machine-learning-based L4 autonomous vehicle system to driverless commercial deployment.
• Direct experience with data generation/curation tools, active learning, and automated edge-case extraction pipelines.

Responsibilities

• Lead the implementation of ISO 21448 (SOTIF) and UL 4600 frameworks across Gatik's AI/ML pipelines.
• Derive safety requirements and performance boundaries (e.g., false positive/negative limits, edge-case coverage metrics) for neural networks and deep learning models.
• Design and validate deterministic AI safety guardrails, out-of-distribution (OOD) detectors, and runtime monitoring mechanisms that fall back to fail-safe actions upon model degradation or uncertainty.
• Establish criteria for training and validation dataset completeness, operational design domain (ODD) coverage, class imbalance, and synthetic vs. real data fidelity.
• Conduct systematic HAZOP, STPA, and SOTIF risk analyses specifically targeting deep learning failure modes (e.g., adversarial examples, sensor degradation, lighting/weather edge cases).
• Author structured safety case arguments (Goulburn/GSN or Goal Structuring Notation) demonstrating acceptable residual risk for AI components in driverless operations.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior