Korean Audio Annotation Specialist — Remote, Part-Time

Crossing Hurdles; Malaysia; Teletrabalho; MYR 570 000 - 1 006 000; Tempo parcial

Resumo da oferta

Crossing Hurdles is seeking a detail-oriented audio annotation specialist to review multilingual datasets for AI evaluation. You will verify transcription accuracy, assess speaker diarization, timestamps and metadata, and ensure strict adherence to annotation guidelines.

Native Korean proficiency with exceptional listening skills is required. Candidates should have verbatim transcription and multi-speaker annotation experience, with strong JSON data familiarity and a linguistics or phonetics

Qualificações

• Experience in audio annotation for AI evaluation.
• Verbatim transcription and multi-speaker annotation experience.
• Strong attention to detail and data accuracy.
• Familiarity with JSON-based data formats.
• Background in linguistics, phonetics, or transcription.
• Ability to identify audio issues like noise and clipping.

Responsabilidades

• Review multilingual audio datasets for AI evaluation.
• Validate audio quality and verify transcription accuracy.
• Assess speaker diarization, timestamps, and metadata elements.
• Ensure strict adherence to annotation guidelines.
• Identify and report subtle audio issues present.
• Evaluate complex conversational dynamics in recordings.
• Maintain detailed documentation of annotation processes.

Conhecimentos

Audio annotation

Transcription accuracy

Multi-speaker tagging

JSON data

Linguistics background

Background noise detection

Formação académica

Linguistics/Phonetics/Transcription background

Descrição da oferta de emprego

Crossing Hurdles is seeking a detail-oriented audio annotation specialist to review multilingual datasets for AI evaluation. You will verify transcription accuracy, assess speaker diarization, timestamps and metadata, and ensure strict adherence to annotation guidelines.

Native Korean proficiency with exceptional listening skills is required. Candidates should have verbatim transcription and multi-speaker annotation experience, with strong JSON data familiarity and a linguistics or phonetics