Robotics ML Engineer: Edge Perception & Real-World Ops

OpenMind; San Francisco (CA); On-site; USD 130,000 - 210,000; Full time

Job summary

OpenMind in San Francisco, CA, is building the operating system for robotics. You will be the engineer turning research into models that run reliably on physical robots, handling training, evaluation, and inference systems behind our models.

You will work with continuous real-world embodied data from humanoid and quadruped robots in public spaces. You will ship models as containerized services in our OM1 stack, and own latency, reliability, and field regressions while collaborating with

Qualifications

• MS or PhD in ML, robotics, CS, or related field.
• 1+ years of industry or applied research experience shipping ML models.
• Strong programming in Python/Go/C++ and PyTorch or JAX.
• Experience with multimodal models, perception, or reinforcement learning and imitation learning.
• Edge deployment on hardware with latency/memory constraints (e.g., NVIDIA Jetson).
• Experience with TensorRT, ONNX, or other inference toolchains.
• Experience with ROS2 and Docker-based deployment on robots.
• Experience building datasets or benchmarks used outside your team.
• Publications at CoRL, RSS, ICRA, NeurIPS, ICML, or ICLR.
• Bias toward models that work in the real world, not only on benchmarks.

Responsibilities

• Build multimodal perception models for social awareness (active speaker detection, person tracking, engagement estimation).
• Scale pipelines turning multi-hour logs into labeled datasets for training and evaluation.
• Ship containerized models in the OM1 deployment stack, managing latency and reliability.
• Collaborate with research, robotics engineers, and deployment teams to productionize models on real robots.
• Integrate, fine-tune, and optimize foundation models for real-time perception on embedded compute.

Skills

ML model training

Production ML systems

Multimodal perception

Education

MS or PhD in machine learning, robotics, computer science, or related field

Tools

Python

Go

C++

PyTorch

JAX

TensorRT

ONNX

ROS2

Docker

NVIDIA Jetson

Job description

OpenMind in San Francisco, CA, is building the operating system for robotics. You will be the engineer turning research into models that run reliably on physical robots, handling training, evaluation, and inference systems behind our models.

You will work with continuous real-world embodied data from humanoid and quadruped robots in public spaces. You will ship models as containerized services in our OM1 stack, and own latency, reliability, and field regressions while collaborating with