What we are writing about.
Field notes from the Scriptix team on speech recognition, custom models, sovereignty, subtitling, translation and digital accessibility.
Speech recognition compared: Scriptix, Speechmatics and HappyScribe
An in-depth procurement-grade comparison of three European-relevant speech-to-text vendors: accuracy methodology, language coverage, custom model mechanics, deployment topology, sovereignty posture, accessibility fit and total cost of ownership.
Vocabulary lists vs custom models: what actually moves accuracy
Most ASR vendors call a glossary a custom model. It is not. Here is why Scriptix does not ship vocabulary lists at all, and only offers real custom models trained on your audio and transcripts.
Digital sovereignty in ASR: where the data actually lives
Where do GoodTape, Speechmatics, HappyScribe, AmberScript and Scriptix actually process your audio? And which of them let you run the engine inside your own walls?
Live subtitling quality: latency, accuracy and the NER model
Word error rate tells you almost nothing about whether a live subtitle is usable. A short guide to the metrics that actually matter on air.
Real time translation in the newsroom: what works in 2026
Cascaded ASR plus machine translation is not new, but the gap between a demo and production is still enormous. Lessons from real newsroom deployments.
The European Accessibility Act, one year in: what changed for media
Twelve months after the EAA came into force, broadcasters and public bodies have a clearer picture of what regulators actually expect from captions, transcripts and translations.