Senior Machine Learning Engineer

Seasats • San Diego, CA

12d • $180,000 - $250,000 • Onsite

About The Position

Seasats' vehicles operate in highly remote and often communications-limited environments. To complete persistent missions at scale, the vehicles must be able to perceive and reason about the world around them without humans in the loop. Seasats has built a reliable and increasingly intelligent autonomy stack incorporating onboard sensors and processing, and machine learning is becoming central to how it evolves. In this role, you'll own the development, training, and edge deployment of the perception models at the heart of that stack. You'll take models from experiment to production: curating fleet data, training and distilling models, optimizing them for embedded hardware, and validating them against real-world maritime conditions. Your work will directly determine how safely and intelligently our vessels navigate dynamic ocean environments. As a senior member of the team, you'll help set the technical direction for onboard ML at Seasats and raise the bar for how models are built, evaluated, and shipped across the company. You'll work independently on experiments while collaborating closely with the vehicle software team to integrate your work into the larger stack.

Requirements

• 7+ years in machine learning, including 5+ years solving perception problems in production systems
• Deep computer vision background, including foundation models and training smaller custom models
• Track record of shipping performant ML models to edge devices and hands-on experience with model optimization for the edge (quantization, pruning, and inference runtimes such as TensorRT or ONNX Runtime)
• Strong proficiency in Python, with expert-level fluency in PyTorch and strong working knowledge of OpenCV
• Experience with MLOps tooling for dataset versioning, experiment tracking, and model deployment
• Experience working with real-time sensor data (e.g., camera, radar, and IMU streams) and familiarity with classical perception techniques (e.g., filtering, tracking, sensor fusion)
• Demonstrated ability to scope projects independently, lead technical efforts, and mentor other engineers
• Strong technical communication skills
• A bias toward getting multiple projects to 80% rather than one project to "perfect", while taking great joy in seeing projects approach perfection over time
• U.S. persons, as defined by U.S. export control laws (U.S. citizens, lawful permanent residents, and individuals granted asylee or refugee status)

Nice To Haves

• Probabilistic algorithms for obstacle and target tracking, and change detection algorithms for a variety of data types
• Real-time multi-modal sensor fusion across LiDAR, radar, camera, and IMU data
• Maritime and/or acoustic data
• Path-planning and control algorithms

Responsibilities

• Help define and drive the ML roadmap for vehicle perception
• Scope and run ML experiments with clearly measurable end states that would positively impact vehicle performance if successful
• Train, fine-tune, and optimize models sized for compute-limited edge hardware, and validate real-time performance on target hardware
• Build and maintain the data pipeline: analyzing, organizing, and labeling datasets from our fleet to drive data-driven development
• Establish evaluation frameworks and metrics that give the team confidence in model behavior before it goes to sea

Benefits

• Stock options
• Competitive insurance (including a 99% employer-covered Gold HMO plan or other options)
• 401k matching up to 4% of salary
• Four free lunches per week
• An employee activity fund
• A pet-friendly office
• Unlimited/Flex PTO

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed