Kelvin D'bruyne

I am a results-driven Global AI Data Specialist with extensive experience supporting North American and European AI firms, focused on bridging raw data and model alignment through adherence to structured guidelines. I specialize in multi-modal datasets — audio, video, and image — and maintain a 99.8% Quality Assurance record while operating from a fully redundant remote setup.

As a Senior AI Training Consultant, I evaluate LLM outputs for truthfulness, helpfulness, and safety, execute complex multi-modal annotation including semantic segmentation, and earned Master Annotator status within three months through consistent 100% QA scores. Previously, I delivered high-quality datasets for Human-in-the-Loop processes, surfaced 50+ critical edge cases that improved guidelines, and managed large-scale audio transcription and diarization under tight SLAs.

I pair advanced technical tools (Python, SQL, Labelbox, CVAT, SageMaker) with precise technical writing and C2-level English to act as a Subject Matter Expert on linguistic nuance. I seek to contribute high-accuracy annotation, rigorous QA, and RLHF expertise to teams building safe, helpful, and well-aligned AI systems.