Galician Audio Transcriber - Expert Linguistic Judgment

perle; Spain (TX); Remote; EUR 1,653,000 - 3,031,000; Full time

Job summary

Perle is seeking expert Galician transcribers to convert native-speaker audio into precise transcripts for AI model training. You will apply Perle's project style, handle speaker labeling, time-align transcripts, and manage disfluencies and non-speech events in real-world audio.

Work involves principled orthography decisions for dialectal and code-switched speech, with QA collaboration and periodic calibration. Flexible scheduling and per-task compensation support varying workloads.

Qualifications

• Native fluency in Galician and deep immersion in Galicia.
• Mastery of Galician orthography per Real Academia Galega guidelines.
• Ability to distinguish Galician from Portuguese and from Spanish in mixed audio.
• Comfortable with Galician-Spanish code-switching as it occurs naturally.
• At least two years of professional transcription, captioning, or linguistic annotation experience.
• Reliable internet, headphones, and a quiet workspace.

Responsibilities

• Transcribe audio into accurate, time-aligned Galician text per style guide.
• Apply conventions for speaker identification, overlapping speech, disfluencies, and non-speech events.
• Make orthographic decisions for dialectal and code-switched speech.
• Flag ambiguous segments and audio quality issues using project conventions.
• Self-review submissions against Perle's quality bar before delivery.
• Engage with QA reviewers on edge cases and style updates.

Skills

Galician fluency

Orthography mastery

Code-switching

Transcription experience

Tools

Annotation platform

Job description

Perle is seeking expert Galician transcribers to convert native-speaker audio into precise transcripts for AI model training. You will apply Perle's project style, handle speaker labeling, time-align transcripts, and manage disfluencies and non-speech events in real-world audio.

Work involves principled orthography decisions for dialectal and code-switched speech, with QA collaboration and periodic calibration. Flexible scheduling and per-task compensation support varying workloads.