Spring ’27 Intern – ML / Perception / Robotics

moss • San Francisco, CA

2d

About The Position

At moss.ag, we build robots to go where humans won't, digitizing the physical outdoor world to make it machine-readable. We are looking for an ML, Perception, or Robotics Intern to join our fast-moving team building systems that allow our robots to understand and operate in complex outdoor environments. Depending on your interests and experience, you may work on 3D perception, multimodal ML, sensor fusion, mapping, localization, navigation, or autonomy. You might train detection and segmentation models on large custom datasets, fuse LiDAR and camera data, build 3D mapping and localization systems, develop robot behaviors, or optimize models and algorithms for real-time deployment on edge GPUs. Our robots operate through changing light, harsh shadows, motion, dust, dense vegetation, uneven terrain, and severe occlusions. You’ll work across the full loop — from data and research to deployment, field failures, and iteration on real robots.

Requirements

• Impressive technical projects beyond the classroom in ML, computer vision, perception, robotics, mapping, or related fields
• Strong programming skills in Python, C++, or Rust
• Hands-on experience with at least one of the following: Machine learning or computer vision, 3D perception, point clouds, or sensor fusion, Localization, mapping, navigation, controls, or robot behaviors
• Comfortable working in Linux and debugging real systems
• Excited to learn quickly, take ownership, and test your work on real robots

Nice To Haves

• Experience with PyTorch, LiDAR, cameras, ROS, CUDA, model deployment, or edge computing

Responsibilities

• Own projects from research and design through deployment and iteration on real robots
• Build and evaluate ML, perception, mapping, and autonomy systems
• Train models on custom outdoor datasets collected by our robots
• Develop sensor fusion across LiDAR, cameras, GPS/IMU, and other sensors
• Build localization, navigation, controls, and robot behaviors
• Develop tools for data collection, labeling, training, evaluation, simulation, and field debugging
• Optimize models and algorithms for latency, memory usage, and reliability on edge hardware
• Test and deploy robotic systems during real-world customer operations
• Collaborate closely with perception, software, electrical, and mechanical engineers

What This Job Offers

• Job Type: Part-time
• Career Level: Intern
• Education Level: No Education Listed