Dayvid The Grapher

Analytical and precision-driven AI Data Quality professional with extensive hands-on experience in AI data annotation, multi-modal dataset validation, and model training workflows across global platforms including Outlier (Aether Project), Atlas Capture, and Telus. Proven expertise in rubric-based evaluation, content policy compliance, and high-volume quality assurance (QA) auditing. Skilled at executing complex project instructions to deliver high-integrity training data for LLM optimization, prompt/response evaluation, and RLHF workflows.

AI Trainer & Data QA Reviewer - Outlier (Aether Project), Atlas Capture, Telus, Lighthouse 3 - Remote

(2025-01)

• Execute detailed data annotation, prompt evaluation, and ranking workflows to optimize LLM performance and support RLHF training objectives.
• Audit and validate multi-modal datasets (text, image, and video), isolating labeling errors and maintaining a consistent 95%+ QA accuracy rating against strict project rubrics.
• Interpret complex, evolving project guidelines to ensure absolute compliance, reducing dataset rework rates by 15% through rigorous edge-case handling.
• Author clear, analytical rationales to justify model response ratings, detailing factual accuracy, safety policy adherence, and hallucination detection.
• Review, capture, and submit high-quality media datasets via Atlas Capture, strictly monitoring environmental framing and lighting specifications to achieve 100% QA compliance.
• Collaborate asynchronously with global teams via Jira, Slack, and Notion to deliver structured feedback reports and streamline remote data pipelines.

Bachelor of Science (B.Sc.) - Biological Sciences - University of Abuja