Shivam Upadhyay

At RMSI, I've progressed from production annotation to QA and Team Lead, leading a 30-person annotation team across more than 100 project assignments.

I've personally annotated 300,000+ images and QA-reviewed 400,000+ images at 99% QA accuracy, working with bounding boxes, polygons, segmentation, keypoints, tracking, classification, and geospatial imagery.

I also completed full-time video annotation work across object tracking, temporal and timestamp annotation, action segmentation, and human activity labeling. I've trained and onboarded 100+ annotators while running guideline, boundary, class, tracking, and edge-case QA workflows.

Through Shift, I annotate and review egocentric robotic manipulation data for VLA model training after being selected for invitation-only QA/QC access. At Handshake AI, I evaluate LLM responses for quality, accuracy, and guideline adherence.