IZUCHUKWU CHINEDU

Evaluated generative AI model responses using detailed rubrics to assess factual accuracy, prompt adherence, safety criteria, and hallucination detection. Completed data labeling work including Named Entity Recognition (NER) and sentiment classification for large-scale multilingual text datasets. Managed a team of 5 annotators by setting quality control guidelines, performing QA audits, and reducing overall error rates by 30%. Developed Python preprocessing scripts to clean, structure, and format dataset inputs, improving pipeline turnaround time by 20%. Managed multilingual dataset quality assurance across 7 European languages (English, French, Dutch, Czech, Polish, Portuguese, Ukrainian). Produced 1,200+ hours of accurate transcriptions and data processing under strict confidentiality.

Formatted, verified, and annotated text and audio data with a 99% client satisfaction rating. Trained new hires on quality control standards, annotation tools, and guideline compliance to help maintain consistent output quality.

Produced high-quality transcriptions for corporate webinars and podcasts, maintaining a 99% client satisfaction rating. Trained new hires on transcription quality standards and the tools used to deliver consistent results. Worked with an emphasis on accuracy, formatting quality, and reliable delivery within client expectations.