Research Engineer/Senior Research Engineer, 3D Auto-Labeling and Data Flywheel

PlusAI•Santa Clara, CA

25d•$130,000 - $220,000

About The Position

PlusAI is hiring a Research Engineer or Senior Research Engineer to develop 3D obstacle auto-labeling methods and data-flywheel capabilities for autonomous trucks. The role converts curated, multi-sensor fleet recordings into accurate 3D object labels using offline models and compute, with the goal of enabling reviewers to verify labels rather than create them manually. The work addresses sensor fusion, temporal consistency through occlusion, and assessing when machine-generated labels are reliable enough. The role initially focuses on 3D obstacles and may expand to other parts of the data flywheel. Level is based on experience, technical depth, responsibility, and demonstrated impact. The engineer will develop and evaluate methods against reviewed ground truth, improve label quality while reducing human review, and build and operate fleet-scale labeling pipelines. Responsibilities also include delivering training-ready labels, working with modeling teams to use model failures and long-tail scenarios to guide future data curation, and following the company's Quality Management System requirements. Senior-level scope includes leading designs and mentoring.

Requirements

BS, MS, or PhD in Computer Science, Robotics, Electrical Engineering, or a related technical field, or equivalent practical experience.
Hands-on experience developing and evaluating 3D perception methods, such as detection, tracking, or sensor fusion, using a framework such as PyTorch.
Solid fundamentals in 3D geometry, coordinate frames, and camera and lidar calibration, including experimentation, error analysis, and principled metrics.
Strong Python skills, including NumPy, Pandas, and SQL, and working C++ skills where performance matters.
Experience running large offline workloads with GPU batch inference.
Knowledge of data-centric machine learning, including how training-data composition affects model behavior, how to evaluate against ground truth, and how to measure and reduce annotation cost.
Self-driven approach, strong sense of responsibility, ability to learn quickly, and ability to take ambiguous problems through to delivery.
For the senior level: experience owning an end-to-end system, producing design documents others build on, and mentoring.

Nice To Haves

Auto-labeling, offline perception, or ground-truth generation for autonomous driving or robotics, including annotation-tool and vendor integrations.
Experience improving models through data, including active learning, data flywheels, hard-example mining, uncertainty or disagreement signals, and dataset curation.
Experience with lidar SLAM, structure-from-motion, factor-graph optimization (GTSAM or Ceres), or stereo and monocular depth estimation.
Experience with foundation and multimodal models, including open-vocabulary detectors, promptable segmentation, vision-language models, embeddings, and vector search.
Heavy-truck experience, including articulated vehicles, trailer state estimation, or multi-platform calibration, or experience with lane and road-geometry labeling pipelines.

Responsibilities

Develop and evaluate methods that turn multi-sensor recordings into accurate 3D object labels offline, and measure results against reviewed ground truth.
Improve label quality and reduce human review by identifying and addressing failures in models, sensor fusion, or data.
Bring state-of-the-art perception and foundation models into offline use and adapt them to the company's vehicles and sensors.
Build and operate fleet-scale labeling pipelines, including batch GPU inference, data extraction and alignment, human-review selection, annotation-tool integration, monitoring, throughput and cost management, and regression testing.
Deliver training-ready labels and work with modeling teams to feed model failures and long-tail scenarios back into data curation and labeling.
At the senior level, lead designs and mentor.
Perform work in accordance with the company's Quality Management System requirements and contribute to continuous improvement.