Isaac Kaiso

I’m a meticulous AI content evaluator and data quality specialist with 3+ years delivering consistent, high-accuracy assessments across multiple AI/ML platforms. I evaluate 1,000+ AI outputs using structured rubrics to surface logical inconsistencies, factual gaps, and instruction-following failures—while maintaining a 95%+ quality rating.

I bring disciplined rubric-based assessment rooted in RLHF / feedback loops, including developing and applying scoring criteria to flag weak reasoning chains, off-prompt responses, and surface-level issues. I also identify recurring model error patterns and summarize findings in weekly Excel quality reports so teams can improve performance without losing throughput.

Beyond evaluation, I support data quality and reporting as a Junior Data Analyst (Excel, Google Sheets, and QA documentation), cleaning 50+ datasets with zero critical errors over 6+ months. I thrive on self-directed remote work with zero missed deadlines, collaborating asynchronously with trackers and shared documents, and delivering audit-ready documentation with clear, structured written feedback.