Founding Robotics Simulation Engineer – Build Real-World Digital Twins

Drift; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Drift is seeking a Founding Robotics Simulations Engineer in San Francisco to help build the simulation infrastructure that accelerates robotics deployment from lab to real world. You will own projects end to end, from early customer conversations to final robot deployment, and work closely with teams building humanoids, mobile manipulators, and industrial robots.

You will create physics-based digital twins, validate against real data, and turn insights into reusable tooling across the Drift

Qualifications

• Strong physics simulation fundamentals underpin robot behavior.
• Hands-on experience with MuJoCo, mjlab, or Isaac Sim.
• Robot modelling using URDF/MJCF/USD.
• Experience in sim-to-real or real-to-sim deployment.
• Willingness to work directly with hardware.

Responsibilities

• Build physics-accurate digital twins of robots, worlds and assets.
• Translate simulation policies to real hardware and identify gaps.
• Design domain-randomized simulation scenarios for customer insights.
• Calibrate sim against real data using system ID and parameter estimation.
• Collaborate with customer engineering teams to scope problems and deliver milestones.
• Develop reusable tools and infrastructure across the Drift platform.

Skills

Physics simulation fundamentals

MuJoCo / mjlab / Isaac Sim

Robot modelling (URDF / MJCF / USD)

Sim2real / real2sim experience

Hardware debugging / hands-on robotics

Tools

ROS2

OpenSim

Unity

Unreal

Docker

Linux

Job description

Drift is seeking a Founding Robotics Simulations Engineer in San Francisco to help build the simulation infrastructure that accelerates robotics deployment from lab to real world. You will own projects end to end, from early customer conversations to final robot deployment, and work closely with teams building humanoids, mobile manipulators, and industrial robots.

You will create physics-based digital twins, validate against real data, and turn insights into reusable tooling across the Drift