Annotator
In my role within the Atlas Capture ecosystem, I focused on high-precision data annotation to support the development of frontier AI models. My core responsibilities involved dense labeling and atomic action annotation of egocentric (first-person) video data, where I meticulously identified and categorized complex hand-object interactions. This work required extreme attention to detail to ensure that every human action was labeled in the correct sequence, maintaining high-volume throughput without sacrificing data integrity. Beyond basic labeling, I performed quality audits on video segments and audio-to-text transcriptions, verifying that AI-generated data aligned perfectly with real-world observations. This experience allowed me to master specialized annotation software and established my ability to manage the rigorous data standards necessary for training advanced machine learning algorithms.