About the Team
Avride is at the forefront of autonomous mobility, developing and deploying cutting-edge self-driving cars and delivery robots. We’re shaping the future of transportation and logistics, and our Data Annotation and Behavior Analysis teams play a vital role in bringing that vision to life. Our work is organized around two core domains: autonomous vehicles and delivery robots. While each presents unique behavioral patterns and annotation challenges, we follow a unified framework for annotation quality, behavior evaluation, and feedback integration. The primary goals of these teams are to deliver high-quality labeled data and to understand and evaluate real-world vehicle behavior, supporting continuous improvement in performance, safety, and user experience.
About the Role
We are looking for a Data Annotation Engineer who will focus on testing and optimizing labeling workflows. You will be responsible for developing and refining annotation pipelines, identifying and resolving issues in 3D annotation tools, and creating metrics to assess data quality.
Your work will directly impact how efficiently our annotation team produces accurate and high-quality data for training machine learning models.
We use Python and ClickHouse, and we’d love to hear your ideas on how to use them effectively.
What You'll Do
Annotation Workflows
- Develop and maintain 3D data labeling pipelines.
- Collaborate with backend and frontend teams to improve interface usability and the data delivery pipeline.
- Work with external vendors to ensure consistent data quality standards.
Testing & Troubleshooting
- Test and evaluate annotation tools to identify bugs, UX issues, and performance bottlenecks.
- Perform various types of testing (functional, regression, exploratory) on new features.
- Work closely with developers and product managers to promptly resolve issues.
Quality & Analytics
- Create and monitor annotation quality metrics, analyze trends, and recommend improvements.
- Document workflows, best practices, and edge cases for 3D point cloud labeling.
- Use Python and ClickHouse to analyze data, monitor performance, and support debugging.
What You'll Need
- A degree in a relevant field (Computer Science, QA, or other technical disciplines);
- Understanding of testing processes (functional, regression, smoke, UI/UX) and experience writing clear bug reports;
- Python skills for data processing and analysis (Pandas, NumPy);
- Ability to query databases (experience with ClickHouse is a plus).
Nice to Have
- Hands-on experience with LiDAR point clouds;
- Hands-on experience in data annotation or data visualization (preferably with 3D datasets);
- Experience using semi-automated labeling tools or active learning methods.
Candidates are required to be authorized to work in the U.S. The employer is not offering relocation sponsorship, and remote work options are not available.
