Felix Leo

Train LLMs, NLP, and vision-language models via image and video annotation and conduct timestamped captioning and multimodal dataset QA to improve contextual accuracy and labeling consistency.

Labeled sensor and 3D motion data for robotic perception and automation models, maintaining annotation consistency and documentation to optimize ML workflows.

Annotated aerial imagery for solar suitability and roof geometry, validated automated labeling outputs, and supported tool testing in CVAT and Supervisely for renewable energy assessment.

Performed multimodal labeling (text, image, audio, video), simulated AI-human conversations for LLM evaluation, and transcribed environmental audio datasets to support diverse AI projects.

Labeled RGB, infrared, and sonar data for maritime navigation and environmental AI models, annotating ships, wildlife, debris, and harbor features with precise multispectral labels.

Mapped site elements and excavation zones from imagery to support construction progress monitoring AI models, delivering consistent annotations for infrastructure and renewable installations.

Pursuing a Bachelor of Education in Technology Education with focus on building construction technology and IT systems, expected completion in December 2025.