About the Role
At Skild AI, we are building the world's first general purpose robotic intelligence that is robust and adapts to unseen scenarios without failing. We believe massive scale through data-driven machine learning is the key to unlocking these capabilities for the widespread deployment of robots within society. Our team consists of individuals with varying levels of experience and backgrounds, from new graduates to domain experts. Relevant industry experience is important, but ultimately less so than your demonstrated abilities and attitude. We are looking for passionate individuals who are eager to explore uncharted waters and contribute to our innovative projects. We are looking for a highly detail-oriented and technically strong Senior Master Annotator to define, maintain, and continuously improve annotation standards for robotic datasets. Various type of data sets can be generated which needs to be annotated and quality checked before they are put in production. This role will act as the gold standard authority for labeling tasks, own Inter-Annotator Agreement (IAA) measurement, train and calibrate annotators, and ensure that annotation quality meets the high standards required for robotics model training. The ideal candidate understands image, video, 3D and audio annotation, quality metrics, and how annotation accuracy impacts model performance.
Responsibilities
- Define and maintain gold-standard annotations and benchmark (“gold set”) datasets for new and existing tasks
- Establish clear acceptance criteria, tolerance thresholds, and task-specific quality standards (e.g., frame-level tolerances)
- Develop, document, version-control, and continuously refine annotation guidelines and SOPs based on disagreement patterns and quality findings
- Design and implement Inter-Annotator Agreement (IAA) frameworks (e.g., Cohen’s/Fleiss’ kappa, tolerance-based agreement models) and produce regular quality reports
- Analyze disagreement trends, identify root causes, and recommend data-driven process improvements
- Lead onboarding and ongoing training for annotators, including structured calibration sessions
- Build and maintain edge-case libraries and example banks to improve annotation consistency
- Provide targeted feedback and retraining to improve annotator performance and alignment with gold standards
- Conduct precision-based QC audits according to task requirements and validate pilot batches prior to production scaling
- Review escalated edge cases and partner with QA and Operations teams to optimize sampling strategies and quality processes
Qualifications
- 4–7+ years of experience in data annotation, quality control, or video labeling environments
- Strong expertise in temporal video annotation; experience with robotics, computer vision, or autonomous systems datasets preferred
- Deep understanding of Inter-Annotator Agreement methodologies (e.g., Cohen’s/Fleiss’ kappa), quality metrics, sampling strategies, and gold set creation
- Experience mentoring, training, or guiding annotators to improve accuracy and consistency
- Strong analytical skills with proficiency in Excel or Google Sheets (Python preferred but not required)
- Ability to translate ambiguous SOPs into precise, actionable annotation guidelines
- Exceptional attention to detail and sound judgment in ambiguous visual tasks
- Data-driven, process-oriented mindset with strong ownership and accountability
- Excellent written and verbal communication skills