Skip to content
Search jobs inside your AI agent. Claude, ChatGPT, Claude Code, Codex, Gemini, Cursor and Grok
Connect

RWS Group

NewBe an early applicant

[TrainAI] Staff Robotics Engineer Technical Lead

Posted Updated

Description

Job Purpose

About the Role

This is the first and most senior technical hire for our China physical AI team. You will build and lead the team in China, own the technical delivery of our capture stack, our data pipeline and our first humanoid deployment in China, and act as the technical counterpart to robotics and embodied AI clients in China.

You will work closely with our research team in the US to decide what data our clients need and how we produce it, and you will own technical direction and delivery in China.

This is a hands-on engineering leadership role, not a blue-sky research position. It combines four things that are usually separate jobs:

  • practical judgment about what makes robot training data useful
  • hands-on capture hardware and data-format engineering
  • team leadership
  • client-facing technical credibility

We are looking for a pragmatic builder who ships working systems, makes sound trade-offs under real deadlines, and stays close to the code and the hardware.

About Train AI

RWS TrainAI is the AI data and evaluation business of RWS Group, working with leading AI developers worldwide. We are building a physical AI practice: egocentric and multi-sensor human-demonstration data, data pipelines and formats designed for robot learning, multilingual and culturally localized interaction for humanoid robots in real venues, and human evaluation of how robots behave around people. Our advantage is a global, multilingual human network and a focus on data that is culturally diverse, consented and auditable.

Job Overview

Key Responsibilities

Lead the China technical team

  • Hire, lead and develop the China engineering team: robot systems integration, robot application software, capture hardware, data pipeline, and robotics research engineering.
  • Set technical standards, review designs and code, make the architecture decisions, and stay hands-on in the code and on the bench.
  • Prioritise ruthlessly: ship what clients need now, and keep the system simple enough to scale.
  • Lead the engineering of our first humanoid deployment in China: platform integration and the application layer (speech, dialogue, operator console, logging) on a commercial humanoid, with interaction design and evaluation coming from our research team in the US.
  • Be the technical counterpart to the humanoid platform maker’s developers: SDK access, feature requests, joint debugging and the technical agenda of the partnership.
  • Work with the programme manager, operations leads and the China sales team so that what engineering builds is what delivery and clients need.

Own the data specification

  • Work backwards from how VLA, world-model and imitation-learning systems are trained to define our data: episode schema, action and state representation, time-sync tolerances, segmentation, and annotation taxonomy.
  • Keep our exports compatible with the formats clients use (e.g. LeRobotDataset, RLDS, MCAP, HDF5), and decide where we go beyond them.
  • Close the gap between human demonstration and robot-usable data, including hand pose and human-to-robot retargeting.
  • Use quick, practical tests (small fine-tuning runs, data ablations) to settle format and quality decisions with evidence rather than opinion.

Build the capture stack

  • Lead the roadmap from phone-based capture with fisheye optics, to wrist-worn sensors and lightweight RGB-D, to UMI-style rigs via partners.
  • Own the calibration protocol and the technical acceptance criteria for any third-party capture hardware.
  • Work with our software engineers on device integration and in-app quality checks, and with the hardware commercial lead on supplier choices.

Own quality and the pipeline

  • Define the quality system: on-device checks, server-side validation, batch QC, and the metrics that go with them (usable-hour yield, inter-annotator agreement, label error rate).
  • Make sure clients' data is stored and processed in the right region, in line with PIPL and data-security requirements.

Be the technical face to clients and partner with Sales in China

  • Lead technical discussions with robotics and embodied AI companies: turn their model requirements into collection specifications, and explain our methods with credibility.
  • Take part in scoping, pilots and technical reviews alongside commercial colleagues.
  • Work closely with our account managers and sales team in China to set technical and engineering priorities, so the team builds what clients will actually buy.
  • Together with our Principal AI Scientist in the US, train the sales and account teams on our physical AI capabilities, so they can position the offering accurately and qualify opportunities.
  • Give Sales realistic roadmaps, timelines and effort estimates for client proposals, and flag early when a request is outside what we can deliver.

Work with research partners

  • Act as the engineering bridge to our academic partner labs and their PhD interns, turning useful research into production practice.

Skills & Experience

Required

  • PhD in robotics, computer science, electrical / mechanical engineering or a related field from a leading university in China or an equivalent top international university.
  • At least 3 years of industry experience in robotics data — collection, curation, labeling or data pipelines — with real-robot or real-world data, not simulation alone.
  • A track record of building and shipping production robotics or data systems used by other teams or customers — hardware, software or pipelines that ran in the real world, not just in papers.
  • Pragmatic engineering judgment: comfortable making trade-offs under deadlines, simplifying, and delivering.
  • Experience leading engineers or researchers, formally or as a technical lead.
  • Solid understanding of how robot learning models are trained — VLA, imitation learning, diffusion policies or world models — and what that means for data.
  • Hands-on experience with multi-sensor data: cameras including fisheye, IMU, depth; calibration; time synchronisation.
  • Experience shipping software on real robots (humanoid, legged or mobile) through a maker’s SDK or ROS 2, with real users: enough to lead and review engineers doing platform integration and robot application work.
  • Strong Python and C++; working knowledge of robot-learning data formats (e.g. LeRobotDataset, RLDS, HDF5).
  • Native-level Mandarin and professional working English, written and spoken.

Preferred

  • Hands-on experience with a commercial humanoid: deploying and evaluating software on the real robot via the maker’s SDK, ROS 2 and NVIDIA Jetson; humanoid foundation-model stacks (e.g. NVIDIA Isaac GR00T or a maker’s open models); or collecting humanoid teleoperation datasets in LeRobot conventions.
  • Experience at an AI data services or physical AI data company in China — data collection, curation, labeling, RLHF / preference data or model evaluation — ideally for robotics, embodied AI or autonomous driving data.
  • Hands-on experience with robot data annotation: task segmentation, action and grasp labeling, success / failure validation, and multi-sensor quality checks.
  • Experience delivering data for clients outside China, or running data collection in overseas markets.
  • Experience with egocentric or human-demonstration data, teleoperation, or UMI-style capture.
  • Experience building data pipelines or datasets used by other teams or customers.
  • Familiarity with robot middleware and logging formats (ROS 2, rosbag / MCAP) — useful for reading client data and for later teleoperation and UMI work.
  • Publications at robotics or ML venues (e.g. RSS, ICRA, CoRL, IROS) — a plus, but shipped systems matter more to us than papers.
  • Experience taking hardware from prototype to small-batch production with Chinese suppliers.
  • Experience working directly with external clients or partners.
  • Familiarity with human-robot interaction or human evaluation of robot behaviour, such as gesture, gaze and body-language generation, speaker identification, or user studies with service robots in public settings.
  • Knowledge of data protection requirements in China (PIPL, Data Security Law).
  • SLAM, navigation, or real-to-sim and simulation work (Isaac Sim-class platforms, scene reconstruction, sim-based evaluation of policies). The team adds a simulation and navigation engineer, and you will set technical direction for that work.

What success looks like

  • First 3 months: core team hired; phone-plus-fisheye capture validated end to end; data spec v1 agreed with at least one client; a live, multilingual humanoid interaction demo in front of an audience, not polished, with an operator in the loop.
  • By 6 months: first client deliveries accepted against agreed quality metrics; calibration protocol documented; wrist / RGB-D prototyping under way; first humanoid deployment running with real people in a China venue.
  • By 12 months: a repeatable, multi-modality capture-to-dataset pipeline running in China; recognised by clients as their technical counterpart; evidence that our data choices improve model results.

Life at RWS

Life at RWS - If you like the idea of working with smart people who are passionate about growing the value of ideas, data and content by making sure organizations are understood, then you’ll love life at RWS.

Our purpose is to unlock global understanding. This means our work fundamentally recognizes the value of every language and culture. So, we celebrate difference, we are inclusive and believe that diversity makes us strong. We want every employee to grow as an individual and excel in their career.

In return, we expect all our people to live by the values that unite us: to partner, putting clients fist and winning together, to pioneer, innovating fearlessly and leading with vision and courage, to progress, aiming high and growing through actions and to deliver, owning the outcome and building trust with our colleagues and clients.

RWS embraces DEI and promotes equal opportunity, we are an Equal Opportunity Employer and prohibit discrimination and harassment of any kind. RWS is committed to the principle of equal employment opportunity for all employees and to providing employees with a work environment free of discrimination and harassment. All employment decisions at RWS are based on business needs, job requirements and individual qualifications, without regard to race, religion, nationality, ethnicity, sex, age, disability, or sexual orientation. RWS will not tolerate discrimination based on any of these characteristics.

RWS Values

Get the 3Ps right – Partner, Pioneer, Progress – and we´ll Deliver together as RWS.

Recruitment Agencies: RWS Holdings PLC does not accept agency resumes. Please do not forward any unsolicited resumes to any RWS employees. Any unsolicited resume received will be treated as the property of RWS and Terms & Conditions associated with the use of such resume will be considered null and void.

Skills

Apply

See also