Labeling Manager

Role Overview

We are seeking a Label Manager to lead the end-to-end data labeling quality process for autonomous driving applications. This role is responsible for defining and enforcing quality gates for labeled datasets across camera, LiDAR, radar, and multi-modal sensor inputs. You will manage external labeling vendors, evaluate new partners, track key KPIs, and work closely with AI, Application, MLOps, and Perception teams to ensure high-quality training data for production-grade machine learning models.

Key Responsibilities

Labeling Quality & Governance

• Define labeling standards, guidelines, and quality gates for all autonomous-driving datasets (2D, 3D, and multi-modal).
• Develop quality metrics and benchmark datasets to continuously evaluate labeling performance.
• Conduct regular audits of labeled data and enforce acceptance/rejection criteria.
• Own the definition of class taxonomies, annotation rules, and edge-case guidelines.

Vendor Management

• Manage relationships with external labeling companies.
• Evaluate, compare, and onboard new labeling vendors based on speed, quality, tooling capabilities, and cost.
• Track vendor-specific KPIs and generate quarterly performance reports.
• Coordinate pilot projects and data annotation trials with new suppliers.

Cross-Functional Collaboration

• Work closely with AI engineers to understand model requirements, failure modes, and dataset gaps.
• Partner with Application engineers to ensure labels align with downstream functional requirements (perception, control, mapping, prediction).
• Collaborate with MLOps teams to integrate labeling workflows into scalable pipelines, tooling, and CI/CD processes.
• Align with hardware and sensor teams regarding new sensor modalities, calibration data, and metadata needed for annotation.

Process & Tooling

• Drive continuous improvement of annotation tools, QA scripts, dashboards, and workflows.
• Work with MLOps to automate quality checks, dataset versioning, and label-diff systems.
• Establish labeling KPIs: accuracy, consistency, speed, error types, coverage, edge-case completeness.
• Own handling of feedback loops between model performance & labeling improvements.

Required Qualifications

• Bachelor’s or Master’s degree in Engineering, Computer Science, Robotics, or related field.
• 3+ years of experience in data labeling, data operations, data quality, or ML dataset management—preferably in autonomous driving.
• Strong understanding of perception tasks (detection, segmentation, tracking, prediction) and annotation types (2D/3D bounding boxes, polygons, tracks, keypoints).
• Experience with sensor data: camera, LiDAR, radar, multi-modal fusion.
• Hands-on with labeling platforms (e.g., Scale, Appen, Sama, CrowdAI, Labelbox).
• Ability to manage external partners and drive KPI-based vendor evaluation.
• Strong analytical thinking and ability to define quality metrics and processes.

Preferred Qualifications

• Experience working with large-scale datasets for autonomous vehicles, robotics, or mapping.
• Familiarity with MLOps pipelines and dataset versioning tools (DVC, ClearML, internal systems).
• Ability to write simple scripts (Python) for data QA or label validation.
• Knowledge of edge-case mining and active learning workflows.

Soft Skills

• Strong communication and coordination across cross-functional teams.
• High attention to detail and strong sense of data quality.
• Vendor negotiation and relationship-management skills.
• Ability to work in a fast-paced, safety-critical domain.