Mach

AI Perception Engineer, Off-Road Autonomy

Ely , United States

About Mach:

Mach builds production-ready autonomy for the world’s toughest off-road environments. Our technology helps OEMs bring autonomous and semi-autonomous heavy equipment to market across industries such as agriculture, construction, mining, defense, landcare, logistics, and maritime.

We are building rugged autonomy systems that operate where conditions are difficult, GPS can be unreliable, visibility can change quickly, and machines must perform safely and consistently in the real world. Our stack combines perception, machine learning, navigation, planning, controls, connectivity, and operator tools to help heavy machines work safer, smarter, and more efficiently.

Position Description:

Mach is looking for an AI Perception Engineer for Off-Road Autonomyto help build the next generation of machine vision and AI systems for off-road autonomous vehicles and heavy equipment.

In this role, you will develop perception capabilities that allow machines to understand complex outdoor environments, detect and classify obstacles, interpret terrain, reason about machine surroundings, and support autonomous decision-making. You will work across the full perception lifecycle, from data collection and labeling strategy to model training, evaluation, deployment, and field validation.

This role will also contribute to emerging vision-language-action model development, helping train and adapt models that connect visual understanding, task context, and machine actions for real-world autonomous systems.

Ideal Experience:

The ideal candidate will have experience in several of the following areas:

• Training and implementing neural networks, including CNNs and other modern deep learning architectures
• Building scalable end-to-end data collection, labeling, training, and evaluation infrastructure
• Implementing optimized edge-deployed machine vision solutions
• Developing machine vision systems for multi-object tracking, filtering, classification, and identification
• Integrating and fusing data from LiDAR, radar, cameras, GPS/INS, and other sensors
• Working with real-world robotic, vehicle, or autonomous systems data
• Debugging perception failures in field environments and improving models based on real-world performance

Requirements for Success:

• Strong experience in computer vision, machine learning, perception, or autonomous systems
• Self-starter who can work independently and manage projects with minimal oversight
• Strong programming skills in a major language such as C++, C, Python, or similar
• Experience with computer vision, robotics, or machine learning tools such as OpenCV, CUDA, TensorRT, ONNX, ROS/ROS2, PyTorch, TensorFlow, or similar tooling
• Ability to build practical systems that balance model accuracy, latency, reliability, and deployment constraints
• Strong debugging, problem-solving, and communication skills

Nice to Have:

• Experience with off-road autonomy, agriculture, construction, mining, defense, maritime, or heavy equipment
• Experience with vision-language models, vision-language-action models, multimodal learning, imitation learning, or embodied AI
• Experience deploying models to embedded or edge compute platforms
• Experience with sensor calibration, 3D perception, SLAM, visual odometry, occupancy mapping, or terrain understanding
• Experience designing data engines, active learning workflows, synthetic data pipelines, or automated labeling systems
• Generous vacation and PTO
• 10+ Paid Holidays
• 75% company sponsored health plan
• Dental, Vision and 401K available
• Equity incentive package

Other Requirements:

• Bachelors Degree in Computer Science
• 3-5 years professional experience in robotics, autonomy, perception, or machine learning, or equivalent hands‑on experience from a rigorous autonomy research lab, graduate program, or competitive robotics/autonomous vehicle team
• US Citizen or permanent green card holder
• Ability to pass a background check

Job Type: Full-time

Salary: $130,000.00 - $180,000.00 per year, dependent on experience and qualifications