Ivan D.

I help AI and technology teams improve the accuracy, consistency, and production-readiness of multilingual speech systems. Working across Dutch, English, and Spanish, I specialize in AI speech data validation, ASR quality evaluation, and structured linguistic QA - ensuring datasets perform reliably in real-world conditions. Unlike task-based transcription providers, I focus on how language data behaves inside models. My work supports teams building and optimizing: • Automatic Speech Recognition (ASR) systems • Multilingual AI training datasets • AI-generated transcript pipelines • Subtitle localization workflows • Language evaluation and benchmarking projects Core expertise includes: • ASR error categorization and dataset quality auditing • AI transcript cleanup with human validation • Multilingual linguistic QA (NL–EN–ES) • Accent and terminology consistency checks • Subtitle timing and localization quality review • Annotation standard enforcement With a background in audio engineering and hands-on AI pipeline support, I understand both the linguistic and acoustic variables that affect model performance. I don’t just deliver transcripts - I help ensure the underlying data is scalable, reliable, and deployment-ready. If you're building, training, or refining multilingual AI systems, I provide structured QA designed for production environments - not one-off tasks.