[Hiring] Asturian Language Specialist – AI Speech Quality Evaluator @Wizwiz AI

Role Description

We are looking for proficient Asturian language speakers to help evaluate the quality of AI-generated speech. You will listen to synthesized audio, compare it with the provided original text, and assess its accuracy, naturalness, and fluency. Your contribution will help improve AI speech synthesis technology for the Asturian language.

• You will evaluate audio clips based on three key dimensions:
• Word Accuracy: Identify and count words that are incorrectly pronounced, added, or omitted.
• Text-to-Speech Accuracy: Rate how accurately the synthesized audio follows the original text on a scale of 1–5.
• Naturalness and Fluency: Evaluate the naturalness, fluency, clarity, voice quality, emotion, and speaking speed on a scale of 1–3.
• You will also provide brief explanations to support your evaluation scores.

Qualifications

• Native Asturian speaker or someone with near-native proficiency in the Asturian language.
• Good listening skills and sensitivity to voice characteristics, including tone, emotion, pronunciation, and speaking speed.
• Strong attention to detail and ability to identify pronunciation and speech errors.
• Previous experience with Text-to-Speech (TTS) evaluation tasks is highly preferred.

Requirements

• Payment: $2.50 per evaluated audio clip.
• Estimated productivity: Approximately 12 clips per hour.
• Initial workload: 100 audio clips for Phase 1.
• Future workload: Additional batches of approximately 100 clips per phase.
• Work schedule: Flexible, with tasks distributed irregularly depending on project needs.
• This is a freelance opportunity, and task availability will depend on project requirements and data distribution.

Benefits

• Compensation: $2.50 per evaluated audio clip, and approximately 12 clips per hour, total estimated earning is $30 per hour.
• Work Arrangement: Remote / Freelance.

Interested?

If you have strong Asturian language proficiency and an interest in AI speech technology, we would love to hear from you! Apply through the registration link to be considered for this project.