Hudson AI, a voice AI company specializing in media content, said on Tuesday it unveiled its voice AI solutions for media production and dubbing, “Hudson Studio”, and an “agentic localization” function at NAB Show 2026, the world’s largest broadcast and media exhibition.

Agentic localization is a workflow in which an agent autonomously performs the entire process of media localization from planning and execution to quality control (QC). The company aims to automate QC, a key bottleneck in the localization process, and reduce human QC time to under 1 hour. It said review of a 1-hour drama often took 2 to 3 days under existing practices.

Hudson Studio supports voice separation, speaker separation, speech-to-text (STT), text-to-speech (TTS) and voice conversion (VC), and supports more than 80 languages. It can separate dialogue from video to create subtitles or synthesize narration and dubbing voices with a desired voice and emotion.

Shin Hyun-jin (신현진), CEO of Hudson AI, said that for media platforms running large-scale localization, even after adopting AI dubbing, much of the budget, quality and schedule issues appear as bottlenecks in QC. He said the company focused on lowering operational barriers by enabling an agent to autonomously carry out everything from project planning and execution to QC.