Native - Speakers Recording Quality Assurance

Upwork; Madrid; Remote; EUR 21,000 - 34,000; Part time

Job summary

Upwork is seeking native Basque speakers to support a Recording Quality Assurance (RecQA) initiative for an AI data provider. You will listen to audio recordings, compare them with the prompt text, and make minor text edits to correct discrepancies.

You will tag major issues using Mismatch, Cutoff, Mumble, or Noise, maintaining high accuracy and consistency. Remote work approx. 7 hours per week with flexible deadlines.

Qualifications

• Native Basque speaker.
• Experience with QA, audio review, transcription, or similar quality-review work preferred.
• Strong attention to detail and ability to identify subtle audio-text mismatches.
• Excellent listening comprehension in the assigned locale.

Responsibilities

• Listen to each audio recording and compare it with the corresponding prompt text.
• Identify subtle differences between spoken audio and written text.
• Correct minor text errors, including missing or added words.
• Identify and correct mispronunciations where applicable.
• Adjust punctuation to reflect pauses and spoken delivery.
• Apply appropriate issue tags for major discrepancies (e.g., Mismatch, Cutoff, Mumble, Noise).
• Follow project guidelines and apply consistent judgment across reviews.

Skills

Attention to detail

Listening comprehension

Quality-review

Judgment calls

Job description

Our enterprise client, a leading provider of high-quality, diverse datasets for AI model development is seeking skilled native speakers to support a Recording Quality Assurance (RecQA) initiative. In this role, you will listen to existing audio recordings, compare them against the corresponding prompt text, correct minor text-audio mismatches, and identify larger recording or quality issues using defined categories. Your careful review will help ensure high-quality speech data for the client's AI training initiative

Work Assignment And Task Summary (What To Expect)

Contributors will evaluate audio recordings against corresponding prompt text. The role requires carefully listening to each recording to make direct text edits for minor errors (e.g., missing/added words, mispronunciations, or punctuation) and applying specific issue tags (such as Mismatch, Cutoff, Mumble, or Noise) for major discrepancies, ensuring high accuracy and consistent quality across all reviews.

Job Description

• Listen carefully to each audio recording and compare it against the corresponding prompt text.
• Identify subtle differences between the spoken audio and written text.
• Correct minor text errors, including missing or added words.
• Identify and correct mispronunciations where applicable.
• Adjust punctuation to accurately reflect pauses, questions, and other aspects of the spoken delivery.
• Apply the appropriate issue tag when the recording contains a significant mismatch or quality problem.
• Distinguish between minor text corrections and issues that require a defined quality tag.
• Follow project guidelines and apply judgment consistently across production units.
• Maintain careful, reliable, and consistent review quality throughout the project.

Qualifications

• Native Basque speaker
• Prior experience with QA, audio review, speech data, transcription, or similar quality-review work is preferred
• Strong attention to detail and ability to identify subtle audio-text mismatches.
• Excellent listening comprehension in the assigned locale.
• Ability to recognize issues such as missing words, added words, mispronunciations, unclear speech, and recording noise.
• Ability to apply defined issue categories accurately and consistently.
• Comfortable making careful judgment calls when reviewing audio recordings.
• Reliable availability for approximately 7 hours per week.

Additional Information

• Work anytime, provided you meet project deadlines and quality standards.
• The work is judgment-heavy, so contributors are expected to prioritize accuracy and consistency rather than simply maximizing speed.
• Tasks and overall workload may shift based on project needs and locale-specific audio volume.
• Consistent availability and high quality are required throughout the assignment.
• Payment is output-based, calculated according to the total number of correctly reviewed utterances.