Saniya Mateen

Results-driven Annotation Quality Analyst with 4.7+ years of specialised experience in LiDAR point cloud annotation, 2D/3D bounding box labeling, sensor fusion datasets, and ADAS ground truth validation. Proven track record of leading cross-functional annotation projects with zero critical defect escapes, improving team throughput by 20%+, and delivering production-grade training data for autonomous vehicle (AV) ML pipelines. Adept with industry-standard tools including Kognic, Rosetta, Morning Star, and Helix.

Consistently recognised as a top performer for annotation accuracy, quality auditing, and team mentorship.

Annotation Quality Analyst - TELUS Digital

(2025-08)

• Orchestrate daily annotation workflows for a team of 10+ annotators across concurrent LiDAR and camera-based AV labeling projects, ensuring on-time delivery with 98%+ quality acceptance rate.
• Conduct structured QA audits on 3D cuboid, semantic segmentation, and polygon annotation tasks — identifying systematic error patterns and implementing corrective feedback loops to reduce rework by ~25%.
• Serve as primary liaison between annotation team and client stakeholders, translating complex annotation guidelines into actionable SOPs, resolving ambiguities that previously caused 15% task rejection rates.
• Track team KPIs (throughput, accuracy, SLA adherence) using QA dashboards; proactively escalate risk items and propose workflow optimisations that improved weekly output by 18%.
• Mentor junior annotators through targeted coaching, reducing onboarding ramp-up from 3 weeks to under 10 days for new hires on LiDAR-based annotation tasks.

Junior Annotator - Intelizign Lifecycle Services Pvt. Ltd.

(2024-06 - 2025-08)

• Executed 2D bounding box, object detection, and gesture recognition annotation tasks on proprietary platforms, achieving quality scores above 95% on client QA reviews.
• Operated LiDAR annotation tools including Kognic, Rosetta, and Morning Star to label point cloud data for pedestrian detection, vehicle classification, and road object segmentation.
• Identified and documented 30+ tool bugs and annotation discrepancies, collaborating with the technical team to improve platform stability for downstream ML model training.
• Supported backend project setup, taxonomy configuration, and dataset structuring — reducing project initialisation time by 20% through standardised configuration templates.
• Delivered onboarding and refresher training sessions on tool usage, annotation guidelines, and quality standards, enabling faster team scaling.

Senior Process Associate – QA - R Process Outsourcing Pvt. Ltd.

(2021-11 - 2024-05)

• Led end-to-end QA validation for 15+ ADAS annotation projects, reviewing LiDAR point clouds, sensor fusion data (2D+3D), and sequential imagery across highway, urban, rural, and expressway scenarios.
• Delivered high-precision outputs for 3D cuboids, semantic segmentation, polygon labeling, and traffic sign recognition, consistently maintaining annotation accuracy above 97% against ground truth benchmarks.
• Identified and resolved annotation inconsistencies in sensor fusion datasets — including misaligned LiDAR-camera overlays and incorrect class labeling — preventing downstream model training errors.
• Established and enforced quality standards, conducting weekly calibration reviews that reduced inter-annotator disagreement by 30% over a 6-month period.
• Handled complex sequential and non-sequential annotation tasks involving multi-object tracking and occlusion handling in dense traffic scenarios, supporting model robustness improvements.
• Streamlined QA workflows in collaboration with project leads, reducing average review cycle time by 22% while maintaining full compliance with client SLA and data accuracy requirements.

Bachelor of Business Administration - BBA - Teresian Degree College, Mysore