Robotics Research Engineer at Abaka AI

<div data-page-id="J3sidwVUMojVHxx3H3YlWosPgyf" data-lark-html-role="root" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-X0MwdwrOOopBVexL4W0lcHJ8gde"> <div data-page-id="RbszdNFNboRjbExNR4XlsiFGg1c" data-lark-html-role="root" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-EmTGd2E5foqF3HxzcqWlOD9lgod"><strong>About Abaka AI</strong></div> <div class="ace-line ace-line old-record-id-HoBXdiDl3odbOkxY1n7lWW6TgHg">Abaka AI is built on one mission: to be the world’s most trusted data partner for AI companies. More than 1,000 industry leaders across Generative AI, Embodied AI, and Automotive AI rely on us to power their data pipelines. With our headquarters in Silicon Valley—and teams in Paris, Singapore, and Tokyo—we support global partners with fast, reliable, and scalable data solutions.</div> <div class="ace-line ace-line old-record-id-AEMWdiMn7o5gWYxmXDilepYEgFe">Our offerings include a diverse catalog of off-the-shelf datasets (image, video, multimodal, reasoning, 3D, and beyond) as well as comprehensive data collection and annotation services. Whether teams need raw data, curated datasets, or full-cycle data engineering, Abaka AI provides the foundation for building high-performance AI systems.</div> <div class="ace-line ace-line old-record-id-QNggdvBWYoHW8qxdjb6lp5IfgWc">&nbsp;</div> <div class="ace-line ace-line old-record-id-WIwpdVhQto0hZ1xHkJKlbg91g76"><strong>About the Role</strong></div> <div class="ace-line ace-line old-record-id-FvX0doe0voxq8sx9ketlCFDwgph">We are looking for a Robotics Research Engineer to join Abaka Robotics Lab, our in-house research group focused on building the data foundation for physical AI.</div> <div class="ace-line ace-line old-record-id-HOUydkaV4odcaExS6relRrtZgNe">We believe the next wave of AI will act in the physical world—and that its greatest bottleneck will be data. Unlike language and vision, there is no internet-scale corpus of robot experience readily available. That data must be captured, synthesized, labeled, and carefully curated. Abaka already works closely with several leading physical AI labs to help build this foundation.</div> <div class="ace-line ace-line old-record-id-BbPzdRXJNoOiIFxD2ZzlUhGvgge">This is a hands-on, research-driven role combining robot learning, data production, and publishable research. You will train and evaluate policies on both simulated and real-world robotic systems, use experimental results to understand which data matters, and design scalable pipelines that turn raw sensor data into high-quality training datasets with minimal human intervention.</div> <div class="ace-line ace-line old-record-id-TAgMdBPwxoGYGKxv1kHl5V3Xgod">You will work with hardware including 6-DoF collaborative robot arms and an ALOHA-style bimanual platform with full teleoperation. Your findings will directly inform the datasets we build for our partners and contribute to publications at leading robotics, computer vision, and machine learning venues.</div> <div class="ace-line ace-line old-record-id-ZypPd24F8o0MQNxcxl7lX58ggMc">&nbsp;</div> <div class="ace-line ace-line old-record-id-HqnkdCyCpoc22VxWg3qlAMsHg1e"><strong>Responsibilities</strong></div> <div class="ace-line ace-line old-record-id-XT5Ud0HfVo6fAHxjcU1lYbk1gIg">Scope 1: Robot Learning Research</div> <ul class="list-bullet1"> <li class="ace-line ace-line old-record-id-Ous3dxR27oelF0xv45wlDxFHgGf" data-list="bullet"> <div>Train and evaluate robot policies (imitation learning, diffusion policies, VLA, RL fine-tuning) on our data and public baselines, in simulation and on real robots.</div> </li> <li class="ace-line ace-line old-record-id-EwJJdpm4FoSsPNxzwd0lElW3g6b" data-list="bullet"> <div>Turn evaluation results into a quality signal for the data: which data helped, which did not, and why.</div> </li> <li class="ace-line ace-line old-record-id-SPjtdiwkrozFZtxbHUjlwYyCgjd" data-list="bullet"> <div>Track the state of the art: reproduce the papers that matter, run current methods on our hardware, and develop findings that hold up into publications at top robotics, vision, or ML venues.</div> </li> <li class="ace-line ace-line old-record-id-HvfpdJckLoTTqgxo49zlbrGGgKb" data-list="bullet"> <div>Maintain the training and evaluation infrastructure: training runs, evaluation harnesses, and the simulation or world-model environments used ahead of hardware tests.</div> </li> </ul> <div class="ace-line ace-line old-record-id-Npt9dwK8coHQlpxgtrIloFGEgKb">&nbsp;</div> <div class="ace-line ace-line old-record-id-S1j7dS3wcoMDeoxng90l1Y5sgne">Scope 2: Data Pipeline: Synthesis, Annotation, Curation</div> <ul class="list-bullet1"> <li class="ace-line ace-line old-record-id-PxSndOubUoaOIpxf08pl29rVgqh" data-list="bullet"> <div>Design pipelines that turn raw capture (egocentric human video, robot logs, teleoperation, 3D scans) into training data with minimal human labeling: synthesize what was not captured, auto-label with models and geometry, and curate what goes into training.</div> </li> <li class="ace-line ace-line old-record-id-Lgf2d0TEfoNpioxRvUslMW8Mghh" data-list="bullet"> <div>Track which data decisions changed policy performance.</div> </li> </ul> <div class="ace-line ace-line old-record-id-Lp1Pdh1UzoTbEexo5WrlplbDgWh">&nbsp;</div> <div class="ace-line ace-line old-record-id-MD5IdQju9oaDWAx6aVYly93cguc"><strong>Minimum Qualifications</strong></div> <ul class="list-bullet1"> <li class="ace-line ace-line old-record-id-TMHwdvH3mof8r7xlTYTljlelghd" data-list="bullet"> <div>MS or PhD in Robotics, Computer Science, Machine Learning, or a related field.</div> </li> <li class="ace-line ace-line old-record-id-WL2cdRPgcoOnncxVykglH6B3gKb" data-list="bullet"> <div>1 - 3 years of hands-on experience with real robot hardware, through research, an internship, or a project.</div> </li> <li class="ace-line ace-line old-record-id-POmzdquFpotd3fxbb2OlbmKVgsb" data-list="bullet"> <div>Strong machine learning fundamentals: optimization, generalization, evaluation methodology, and end-to-end model training.</div> </li> <li class="ace-line ace-line old-record-id-MYkcdnno9o8Pp9x8FRVlUlQkgih" data-list="bullet"> <div>Strong robotics fundamentals: coordinate frames, kinematics, basic control.</div> </li> <li class="ace-line ace-line old-record-id-IDNndcKGEofQhsxKwBalaEiKgZe" data-list="bullet"> <div>Strong 3D vision fundamentals: camera models, multi-view geometry, point clouds.</div> </li> <li class="ace-line ace-line old-record-id-F5EKd0PjtoLoeTxvd1rlH6GSg38" data-list="bullet"> <div>Rigor in evaluation: claims about a policy or a dataset are backed by measurements.</div> </li> </ul> <div class="ace-line ace-line old-record-id-BgMVdNxOOo0nOnxJHbRl96zxg2A">&nbsp;</div> <div class="ace-line ace-line old-record-id-LqVZdPlCponVoexreAJlDoQLgOb"><strong>Preferred Qualifications</strong></div> <ul class="list-bullet1"> <li class="ace-line ace-line old-record-id-Tjccddqu7oT2N2xXVfBlMP74gnd" data-list="bullet"> <div>Publications or workshop papers at top-tier robotics (CoRL, RSS, ICRA, IROS), vision (CVPR, ICCV, ECCV, 3DV), or ML (NeurIPS, ICML, ICLR) venues. A widely used open-source repository or a dataset adopted by other teams is valued equally.</div> </li> <li class="ace-line ace-line old-record-id-LXBGdDZ1vo7fhFx5pFGl3C0jg6d" data-list="bullet"> <div>Experience with imitation learning or VLA codebases such as LeRobot, ACT, Diffusion Policy, OpenVLA, or π0.</div> </li> <li class="ace-line ace-line old-record-id-TweSd929doIqvyxkMi2lwx0rgtg" data-list="bullet"> <div>Experience deploying a learned policy on a real robot and running a rigorous evaluation.</div> </li> <li class="ace-line ace-line old-record-id-Iz30devJTomKi3xRA8ol6tI2gHy" data-list="bullet"> <div>Practical 3D reconstruction experience: COLMAP, SLAM, 3DGS/NeRF, point-cloud registration.</div> </li> <li class="ace-line ace-line old-record-id-YIH1dsFNwo5AuCxNyfKlwGgQg0f" data-list="bullet"> <div>World models or video generative models, particularly as simulators or data generators.</div> </li> <li class="ace-line ace-line old-record-id-NFmRdwzDVoG9jjxKUn6lHNFKgHh" data-list="bullet"> <div>RL fine-tuning of learned policies, or model-based RL.</div> </li> <li class="ace-line ace-line old-record-id-B1EIdvlJSoge1Ax4b6XlTyVdgEc" data-list="bullet"> <div>Hand and object pose estimation, hand-object interaction, or human-to-robot motion retargeting.</div> </li> <li class="ace-line ace-line old-record-id-HLrAddqLOooHnTxhLOklwNXWggf" data-list="bullet"> <div>Depth in a simulator (MuJoCo, Isaac Sim/Lab, Genesis, SAPIEN), including environment and asset authoring.</div> </li> <li class="ace-line ace-line old-record-id-BAdbdgklFoM9jSxoI7VlUkC5gQc" data-list="bullet"> <div>Domain randomization, procedural scene generation, or sim-to-real transfer.</div> </li> <li class="ace-line ace-line old-record-id-AleadjrM6oMVw4xj6zGlkuBOgre" data-list="bullet"> <div>Teleoperation or data collection systems in the ALOHA / UMI / hand-tracking lineage.</div> </li> <li class="ace-line ace-line old-record-id-OIOddGeNUoUn2wxk1jOlJV9ug8b" data-list="bullet"> <div>Familiarity with egocentric datasets such as Ego-Exo4D, EgoMimic, HOT3D, Project Aria, and DexCap.</div> </li> <li class="ace-line ace-line old-record-id-GooPdoTR7o0lHZxDW2QlxnZJgrh" data-list="bullet"> <div>Dexterous or bimanual manipulation research.</div> </li> <li class="ace-line ace-line old-record-id-CPT6dDrtWoswDaxj3LtljkbGg4e" data-list="bullet"> <div>Camera calibration and multi-sensor time synchronization.</div> </li> <li class="ace-line ace-line old-record-id-CjZ2d2yQGowGoMxxmbPlEWRugLi" data-list="bullet"> <div>Vision-language models applied to robotics: task planning, language-conditioned policies, reward or success detection.</div> </li> <li class="ace-line ace-line old-record-id-KO4jdGNT2oMccMxg8iFld7H1gcg" data-list="bullet"> <div>Auto-labeling or ground-truth generation over real sensor data, with quality metrics attached.</div> </li> <li class="ace-line ace-line old-record-id-FOYQdAJ2zoxFE8xjrzYlxve1gkf" data-list="bullet"> <div>Tactile, force-torque, or other multimodal sensing for manipulation.</div> </li> <li class="ace-line ace-line old-record-id-Jx2DdhfsKoHaXExPk5Vl8mGdgeh" data-list="bullet"> <div>Data curation and dataset-scaling studies: deduplication, mixture ratios, difficulty or diversity scoring, and ablations that identify which data mattered.</div> </li> </ul> <div class="ace-line ace-line old-record-id-MWoMdNtQKoF95LxyZxml0awlg8c">&nbsp;</div> <div class="ace-line ace-line old-record-id-BmxEdJ0WbonfUmxTVw6lsGAWgGj">We recognize that no candidate will match every preferred qualification. If you meet the core requirements and have meaningful depth in several of these areas, we encourage you to apply.</div> <div class="ace-line ace-line old-record-id-MCZrdi62mosdDGx4T1GlO17LgTf">&nbsp;</div> <div class="ace-line ace-line old-record-id-ZItVdjCqoorgKBxgQyolJ6btgsf"><strong>Compensation &amp; Benefits</strong></div> <div class="ace-line ace-line old-record-id-Mx6KdvXhNo3eKQxlIrZl031WgHm">The base salary range for this position is $100,000 - $140,000 USD annually.</div> <div class="ace-line ace-line old-record-id-FbbgdDz2voVSacxDbdglMRWQgeh">Compensation may vary outside of this range depending on a number of factors, including a candidate’s qualifications, skills, competencies, and experience. Base pay is one part of the Total Package that is provided to compensate and recognize employees for their work at Abaka AI. This role is eligible for equity, as well as a comprehensive benefits package (health, dental, vision, PTO, flexible work schedule).</div> </div> </div> </div>