Lead, Data Annotation Team (AI Engineer)
Image
ClassificationI led the Data Annotation team to create high-quality labeled image datasets, directly improving AI model training and production accuracy. I designed and developed a custom internal platform to streamline labeling workflows, implementing validation logic for consistency. Additionally, I built an automated annotation system using llama.cpp to enable efficient on-prem AI-assisted labeling. • Managed large-scale manual and automated image annotation, focusing on object classification and staff identification • Developed modular data annotation workflows tailored for Vision-Language model (VLM) projects on edge devices • Implemented internal validation processes for labeled datasets, resulting in reduced manual review effort and improved accuracy • Coordinated dataset curation for use in person re-identification, segmentation, and monitoring models on NVIDIA DeepStream pipelines