Senior Computer Vision Engineer

SR2; United States; Remote; USD 150,000 - 190,000; Full time

Job summary

SR2 is a robotics startup building perception systems that work outside the lab. This is a remote role on a team that prioritizes strong work, regardless of location.

You will own and ship vision models for robotic understanding, with end-to-end ML lifecycle ownership and real-time edge deployment on embedded hardware. We seek a senior engineer who can independently drive ambiguous problems, mentor others, and deliver robust production-ready software using PyTorch, Python, and C++.

Qualifications

• Senior-level experience shipping computer vision into production.
• Strong PyTorch, Python, and C++ skills.
• Experience with edge deployment and model optimisation (TensorRT or quantisation).
• Ability to work independently on ambiguous problems and communicate clearly.

Responsibilities

• Lead development of vision models for robots to understand the world.
• Own ML lifecycle: data curation, experiments, training, evaluation, deployment.
• Optimise models for real-time edge deployment on constrained hardware.
• Mentor peers through design reviews and guidance.

Skills

Production CV

PyTorch

Python

C++

Model optimisation

Job description

I’m working with a VC-backed robotics company building perception that has to work outside the lab. They’re a small team with a big technical problem, and they want a senior engineer who wants real ownership.

This is a remote role on a team that puts the best work first, wherever you’re based.

What you’ll do

• Lead development of vision models that give robots a reliable understanding of the world
• Own the full ML lifecycle: data curation, experiment design, training, evaluation and production deployment
• Optimise models to run in real time on constrained, embedded hardware
• Raise the bar for the engineers around you through design reviews and mentoring

What you bring

• Senior-level experience shipping computer vision into production
• Strong PyTorch, plus solid Python and C++
• Experience with model optimisation (TensorRT, quantisation or similar) for edge deployment
• The independence to run with ambiguous problems, and clear communication to match