Role summary: Independently checks completed annotations against Valeo's guidelines and acceptance criteria before delivery to Valeo. Owns defect logging, rework verification and quality metrics for the team.
Work location | Onsite at Valeo's Chennai premises. |
Working pattern | 5 days per week, Valeo site calendar and hours |
Key responsibilities
• Validate completed scenes against Valeo's annotation guidelines and acceptance criteria, using full inspection or the agreed sampling plan.
• Record defects by type and severity (Critical, Major, Minor, Observation). Send work back for rework and verify each fix before closure.
• Ensure no scene is delivered with open Critical or Major defects.
• Monitor quality trends by engineer and defect type and support root-cause analysis and targeted coaching.
• Report quality metrics (first-pass acceptance rate, rework rate, defect density and rework closure time) to the Technical Lead each day and week.
• Take part in calibration on gold scenes and maintain the guideline FAQ and clarification log with the Technical Lead.
• Act on feedback from Valeo's acceptance checks and track it to closure.
• Raise guideline gaps or ambiguities as observations for Valeo to clarify.
• 6–7 years in QA, testing or data quality, including at least 2 years reviewing or QC-ing annotated data for computer vision.
• Strong eye for detail on 2D image and/or 3D point-cloud annotations: boxes and cuboids, attributes, class labels, track IDs across frames.
• Hands-on experience with defect tracking, severity classification and rework workflows.
• Good command of Excel for quality trackers and metric reporting.
• Clear written communication for defect feedback.
Good to have
• Automotive ADAS or autonomous-driving perception datasets (camera, LiDAR, sensor fusion).
• Experience with Xtreme1 or similar annotation tools and their review or QC features.
• Knowledge of statistical sampling and quality-gate approaches.
• Experience as an annotator before moving into QC.