Minority language broadcasters get treated as an afterthought by mainstream captioning tools. Generic engines flatten grammatical mutations, miss official terminology, and route live audio to STT systems that barely support the language. Teams end up fixing the same AI mistakes manually, episode after episode.
Every transcription route, translation prompt, and glossary pipeline is aware of Welsh mutations, Irish lenition and eclipsis, Gaelic lenition, Breton consonant systems, Catalan morphology, and Basque case structures. We pin Whisper and Google STT to the right ISO code per language and fall back to Whisper batch where live STT coverage is weak.
Welsh, Irish, Gaelic, Breton, Catalan, or Basque
Whisper or Google STT, pinned to the right code
BydTermCymru, Tearma.ie, TERMCAT, UZEI, Stor-Data and TermOfis packs
Mutation preservation built into the prompt
SRT, VTT, EBU-TT, Premiere XML
"One platform for every Celtic and minority language broadcaster - from S4C to TG4 to BBC Alba to EITB to TV3."
BydTermCymru (Welsh) is live. Tearma.ie, TERMCAT, UZEI, Stor-Data and TermOfis are listed in the SuperAdmin panel as pending - partnership outreach in progress. Want to accelerate a specific language? Get in touch.
"AI supports the language. Humans protect it."