Support remote speech AI projects by evaluating Korean-language voice systems and validating linguistically sound datasets.
What you'll do
- Assess AI-generated speech for pronunciation, naturalness, prosody, emotional register and linguistic quality.
- Identify transcription errors and phonological, morphological and contextual issues in speech recognition outputs.
- Analyze dialects, accents and regional or sociolinguistic variations, and document findings to support dataset diversity.
- Review voice recordings and annotate speech data using phonetic transcription, prosodic markup or morphological tags where applicable.
- Provide structured feedback and quality assurance for speech recognition, text-to-speech, voice generation, voice similarity and conversational AI projects.
Requirements
- A bachelor's degree in linguistics, phonetics, phonology, speech-language pathology, computational linguistics, language sciences or a related discipline; or at least two years of professional experience in linguistic analysis, speech science, language documentation, dialect research, computational linguistics, speech dataset creation or voice AI linguistics.
- Advanced knowledge of Korean phonetics, phonology, morphology and syntax.
- Working knowledge of phonetic transcription, spectrogram and prosodic analysis, acoustic phonetics and language quality assessment.
- Familiarity with linguistic annotation tools and practices, such as IPA, SAMPA or similar systems.
- Ability to apply linguistic frameworks to evaluate speech quality and language naturalness.