Roboticist, Physical AI

Sieve•San Francisco, CA

9h•Onsite

About The Position

Sieve is hiring its first roboticist to own the physical systems used to generate physical AI training data, including robots, capture rigs, and teleoperation stations. The role takes systems from component-level assembly through deployment at partner sites, ensuring they record clean, synchronized data. Work includes hands-on hardware integration and troubleshooting, sensor bring-up and calibration, software development, and maintaining reliable systems in real-world deployments. The position will establish the hardware architecture, interfaces, test standards, and engineering practices for Sieve’s physical AI team. It is based onsite at Sieve’s San Francisco headquarters five days per week, with travel to partner sites.

Requirements

At least 3 years of experience building robotic, mechatronic, or controls systems; a background in robotics, electrical engineering, mechanical engineering, or a similar field.
Experience deploying systems beyond the lab, such as in a factory, warehouse, field site, or customer site, and examples of failures and resulting changes.
Hands-on robot bring-up and integration, with troubleshooting ability across firmware, wiring, and application code.
Workbench skills including soldering, building and repairing harnesses, tracking down power faults, and debugging CAN, serial, USB, and Ethernet links.
Production-quality Python and C++ for hardware running outside the lab; fluency in Linux and ROS/ROS2.
Disciplined engineering practices, including test plans, documentation for shipped work, and measurements to support claims.
Ability to context-switch at a moment’s notice.
Ability to work in person at the San Francisco headquarters and travel to partner sites.

Nice To Haves

Controls experience in areas such as motor control, impedance or force control, trajectory generation, or state estimation.
Experience building teleoperation or data collection setups whose output was used to train robot policies.
Experience calibrating multi-camera systems, synchronizing heterogeneous sensors, or fusing sensor outputs.
Experience with CAD tools such as SolidWorks or Onshape and PCB design tools such as KiCad or Altium.
Experience as an early hire at a startup.
Rust experience.

Responsibilities

Design and bring up capture hardware, including multi-camera arrays, egocentric cameras, depth, IMU, and force sensor packages, and teleoperation stations such as leader-follower arms.
Own camera intrinsics and extrinsics, hardware triggering, and clock alignment across cameras, IMUs, and encoders.
Diagnose and fix system failures such as motor stalls, dropped camera frames, power-rail issues, and communication faults, addressing root causes so failures do not recur.
Set up and operate systems at partner sites, making them durable for daily use, accessible to operators without robotics backgrounds, and repairable onsite.
Write and maintain drivers, ROS/ROS2 integration, recording and logging software, and capture-time QA checks that connect data to Sieve’s pipelines.
Create bills of materials, assembly documentation, validation plans, and standard operating procedures for repeatable deployments.
Establish hardware architecture, interfaces, test standards, and engineering culture for the physical AI team.