Member of Technical Staff, Robotics

As Sieve's first roboticist, you'll own the physical side of our physical AI data: the robots, capture rigs, and teleoperation stations that generate the datasets we deliver to frontier labs. You'll take each system from parts on a table to a setup that records clean, synchronized data at a partner site, and you'll handle both the hardware and the software along the way.

Most of your week will be spent with real equipment: wiring and rewiring, bringing up new sensors, chasing calibration drift, and fixing whatever stopped recording overnight. We want someone who has run systems in real deployments and brings discipline to the work. When a rig works once, you want to know why, and you want it to work the same way at every site. You don't consider a problem closed until the data coming off the system is right.

• Capture rigs: Design and bring up the hardware we record with: multi-camera arrays, head- and wrist-mounted egocentric cameras, depth/IMU/force sensor packages, and teleoperation stations such as leader-follower arms.
• Calibration and time sync: Own camera intrinsics and extrinsics, hardware triggering, and clock alignment across cameras, IMUs, and encoders. If the streams don't line up, the dataset isn't usable.
• Root-cause failures: When a motor stalls, a camera drops frames, a power rail sags, or a bus goes quiet, find the actual cause and fix it so it stays fixed.
• Field deployments: Set up and operate systems at partner sites. Build them to survive daily use, to be run by operators without robotics backgrounds, and to be repaired quickly on site.
• Robot and capture software: Write and maintain drivers, ROS/ROS2 integration, recording and logging, and QA checks that run at capture time and hand data straight to Sieve's pipelines.
• Repeatable builds: Produce BOMs, assembly docs, validation plans, and SOPs so we can stand up the same setup at the next ten sites without you there.
• Team foundations: As the first hire in this area, set the hardware architecture, interfaces, test standards, and engineering culture for the physical AI team.

• 3+ years of experience building robotic, mechatronic, or controls systems (background in robotics, EE, ME, or similar)
• Systems you've deployed beyond the lab (in a factory, warehouse, field site, or customer site), plus stories about what failed and what you changed
• Hands-on bring-up and integration of robots, and comfortable troubleshooting anywhere from firmware to wiring to application code
• Comfortable at a workbench: soldering, building and repairing harnesses, tracking down power faults, and debugging CAN, serial, USB, and Ethernet links
• Production-quality Python and C++ for hardware that runs outside the lab (Rust welcome); fluent in Linux and ROS/ROS2
• Disciplined engineering habits: test plans before builds, documentation for what you ship, and measurements to back up claims
• Able to context switch at a moment's notice
• In-person at our SF HQ [+ travel to partner sites: % TBD]

• Controls background: motor control, impedance/force control, trajectory generation, or state estimation
• Built teleoperation or data collection setups whose output was used to train robot policies
• Calibrating multi-camera systems, synchronizing heterogeneous sensors, or fusing their outputs
• CAD (SolidWorks, Onshape) and PCB design (KiCad, Altium)
• Experience as an early hire at a startup

• 401(k) and full health insurance
• Breakfast, lunch, and dinner covered, plus your choice of snacks
• Ubers covered home

*All roles at Sieve require you to be onsite in San Francisco 5 days per week

Sieve is a multi-modal lab curating the world's highest-quality training datasets — spanning video, audio, images, text, and 3D. We combine exabyte-scale data infrastructure and novel multimodal understanding techniques that push the frontier of foundation models. Video alone makes up 80% of internet traffic, and across modalities, data has become the enabling medium powering creativity, communication, gaming, AR/VR, and robotics. Sieve exists to solve the biggest bottleneck in the growth of these applications: high-quality training data.

We partner with top AI labs and did $XXM last quarter alone, as a team of ~30 people. We also raised our Series A from Tier 1 firms such as Matrix Partners https://matrix.vc/, Swift Ventures https://www.swift.vc/, Y Combinator https://www.ycombinator.com/, and AI Grant https://aigrant.com/.