SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Arctic Engines, part of the Apna Group (a unicorn-backed enterprise AI data operations company), is seeking a QC Lead to own annotation quality for egocentric industrial video datasets. You will define review standards, lead the QC team, identify recurring errors, and ensure delivered annotations meet project requirements.
Key Responsibilities:
- Lead reviewers and establish calibration, review, feedback, and rework processes
- Audit video chunking, temporal action boundaries, keypoint annotations, action labels, and natural language descriptions
- Check timestamp accuracy, coverage, label consistency, and correctness of descriptions against video content
- Define QC checklists and sampling plans; track error rates, reviewer agreement, rejection trends, and quality improvements
- Resolve ambiguous cases, update guidelines, and coach annotators and reviewers
- Validate structured outputs and work with tooling teams to address workflow or export issues
About Arctic Engines:
Arctic Engines is an enterprise-grade AI human data operations company specializing in high-quality training data, RLHF, and human feedback pipelines for frontier AI models. The company has native access to Apna's 60M+ workforce and delivers high-quality training data at scale and speed. Backed by Lightspeed, Tiger Global, Insight Partners, Peak XV, and others.
Requirements:
- 3+ years in video annotation quality assurance, including team leadership
- Direct experience with industrial video, robotics, or Physical AI datasets (mandatory)
- Hands-on expertise in egocentric video annotation, temporal action segmentation, keypoint annotation, action taxonomies, and timestamped descriptions
- Experience leading annotation QC teams and creating clear guidelines and calibration examples
- Ability to analyze errors, run root-cause reviews, and turn findings into corrective action
- Familiarity with video annotation tools and structured outputs such as JSON or CSV
Employment: Full-time, immediate joiners preferred. Work from office in Domlur, Bengaluru (6 days per week).
