AI Voice Quality Analyst — Flexible 15–30 hrs/wk

ixolabs.ai; France; À distance; EUR 35 000 - 52 000; Plein temps

Résumé du poste

ixolabs.ai is seeking evaluators to assess AI-generated English speech for naturalness and fluency, with attention to phonetic rules and prosody. You will identify mispronunciations and robotic cadences, and provide detailed feedback to improve voice models.

The role includes labeling intonation and emotion markers and participating in A/B testing. Strong background in linguistics or phonetics and experience with TTS/ASR are highly valued.

Qualifications

• Native English fluency with an excellent ear for linguistic nuances and speech patterns.
• Background in linguistics, phonetics, speech therapy, or a related field.
• Experience with speech synthesis (TTS) or automatic speech recognition (ASR) systems is highly desirable.
• Proficiency in using evaluation tools for audio data (e.g., Praat, proprietary platforms).
• Strong analytical skills and ability to provide objective, data-driven feedback.
• Detail-oriented and capable of maintaining consistent evaluation standards.

Responsabilités

• Evaluate AI-generated English speech for naturalness, fluency, and adherence to phonetic rules.
• Identify and categorize issues such as unnatural prosody, mispronunciations, or robotic cadence.
• Provide detailed, constructive feedback on AI voice outputs, highlighting areas for model improvement.
• Evaluate speech data with precise labels indicating intonation, stress, and emotional markers.
• Participate in A/B testing of different AI voice models and provide comparative analysis.
• Contribute to the development and refinement of evaluation rubrics and guidelines.

Connaissances

Native English fluency

Linguistics background

Analytical skills

Detail-oriented

Formation

Linguistics/Phonetics background

Outils

Praat

Proprietary evaluation platforms

Description du poste

ixolabs.ai is seeking evaluators to assess AI-generated English speech for naturalness and fluency, with attention to phonetic rules and prosody. You will identify mispronunciations and robotic cadences, and provide detailed feedback to improve voice models.

The role includes labeling intonation and emotion markers and participating in A/B testing. Strong background in linguistics or phonetics and experience with TTS/ASR are highly valued.