Yoshiyuki Nakagawa

I’m a Data Support Specialist who focuses on data quality assurance and AI model evaluation, with a strong track record in prompt engineering and training data generation. I bring a detail-oriented approach to validating datasets, assessing AI responses, and enforcing consistent labeling standards.

Since 2023, I’ve designed and evaluated 1,500+ AI model responses for a Japanese LLM project. I improved response quality scores by 15% through systematic feedback and error analysis, and I built reusable prompt templates that increased content generation efficiency by 25% across client projects.

I also create quality assurance frameworks for data labeling teams, ensuring consistency and accuracy in training datasets. Working with clients, I refine AI outputs based on domain-specific requirements, including finance, marketing, and customer support scenarios, using human-in-the-loop workflows.

Previously (2022–2023), I supported data migration and validation efforts, achieving 99.9% accuracy in transferred records. I generated weekly reporting in Excel/Google Sheets and resolved data-related discrepancies by responding to internal team inquiries, helping stakeholders make decisions with clear, trustworthy metrics.