Thai speech data

Thai-English Code-Switching Speech Data

Thai-English code-switching is common in everyday speech. Here is how to collect and transcribe mixed-language Thai speech data for ASR and conversational AI.

Code-switching is the norm, not an edge case

Thai speakers routinely insert English words for technology, brands, business terms and abbreviations. Monolingual models trained on clean Thai often mis-recognise exactly the terms that matter commercially.

Transcription conventions

  • ▸Whether English words appear in Latin script or Thai transliteration
  • ▸How abbreviations and acronyms are written
  • ▸Language tags at token or segment level
  • ▸Handling of hybrid or partially inflected forms

Collection approach

Bilingual speakers and topic guides drawn from real domains — office, tech support, shopping — produce natural switching more reliably than scripts written to force it.

Tell us what data you need

Send your dataset requirements — language, speakers, hours, environment, transcription and licensing scope. If the specification is still open, our team can help define it.