Audio AI Specialist (Speech Data, Quality & Annotation) - Korean or Japanese

• Review speech and voice recordings for technical, linguistic, and task-specific quality.
• Identify background noise, clipping, distortion, echo, plosives, low signal levels, and unnatural silence.
• Perform speech annotation and quality assurance, including transcription, timestamp validation, voice activity detection (VAD), segmentation, and speaker diarization.
• Verify transcript accuracy, pronunciation, speaker labels, and associated metadata.
• Record high-quality speech based on provided scripts and recording guidelines.
• Apply established audio quality standards and document issues requiring correction.
• Maintain consistency across audio files, transcripts, annotations, and metadata.
• Help improve annotation guidelines, review checklists, and standard operating procedures.
• Collaborate with research, machine learning, and operations teams to meet dataset requirements.

This role focuses on improving the quality, consistency, and usability of speech datasets used to train next-generation AI voice models. The ideal candidate has hands-on experience with audio AI training data, speech annotation, transcription, and audio quality evaluation, along with professional fluency in Korean or Japanese.

Candidates should have a strong ear for audio quality, excellent attention to detail, and the ability to identify subtle issues in real and synthetic speech., * Professional fluency in Korean or Japanese.

• Direct experience working with audio AI training datasets or speech AI evaluation workflows.
• Hands-on experience with ASR, text-to-speech (TTS), speech transcription, speech-to-speech, or conversational voice AI.
• Experience with speech annotation, timestamp validation, segmentation, or diarization.
• Strong ability to identify audio quality issues through listening and audio inspection.
• Experience reviewing transcripts and checking linguistic accuracy and naturalness.
• Familiarity with structured data formats such as spreadsheets, CSV, or JSON.
• Excellent attention to detail and written communication skills.
• Ability to work independently in a remote environment.

Preferred Qualifications

• Experience with Audacity, Praat, ELAN, or similar audio inspection tools.
• Basic Python, Bash, or SQL skills for data quality checks.
• Background in linguistics, phonetics, speech research, localization, or voice recording.
• Experience evaluating both human-recorded and synthetic speech.
• Familiarity with accent and dialect variation.
• Access to a quiet recording environment and suitable recording equipment.
• Understanding of consent, licensing, and compliant handling of voice data.