Colton Leakey

I am an expert dialogue editor with over four years preparing, cleaning, and optimizing spoken-word audio for AI training datasets and broadcast delivery. I excel in fast-paced, high-stakes environments where critical listening and technical judgment are essential.

At OpenAI, I performed repetitive quality control on a 10,000+ hour multilingual speech dataset, improving clean audio classification consistency by 40% through defensible restoration choices. Previously, I managed end-to-end post-production workflows for 75,000+ hours at Appen, achieving a 99.8% technical compliance rate.

I am proficient with iZotope RX, Avid Pro Tools, spectral analysis, and studio-grade monitoring, and I identify and resolve issues like phase problems, noise masking, plosives, and sibilance at scale. I also contribute to AI data workflows through QC metric development, qualitative/quantitative audio reporting, and prompt engineering support.

I bring rigorous quality control, strong time management, clear technical communication, and a collaborative mindset to projects, consistently delivering broadcast-ready dialogue and reducing revision cycles through training and process improvements.