Expert Audio Transcriber, Portuguese (Brazilian)
Core
Transform spoken audio into highly accurate, conventionally formatted written transcripts to support the training and evaluation of speech recognition and conversational AI systems.
Role type
Expert Audio Transcriber (Portuguese)
Builds
Training data and evaluation systems for advanced speech recognition and conversational AI
Domain
AI / Speech Technology / Linguistics
Deliverable
production ML models
Required skills
Native-level Brazilian Portuguese fluency, expert orthographic judgment, regional accent recognition, code-switching handling, speaker identification, disfluency annotation, QA review, web-based annotation platforms, variable-speed audio playback
Preferred skills
Experience with court reporting or broadcasting, familiarity with Acordo Ortográfico
Technologies
Web-based annotation platforms, variable-speed audio playback tools
Responsibilities
Transcribe audio recordings into accurate, time-aligned written text; Apply conventions for speaker identification, overlapping speech, and non-speech events; Make consistent orthographic decisions for dialectal and code-switched speech; Identify and flag ambiguous audio segments; Review completed transcriptions against quality standards; Participate in QA reviews and calibration sessions; Incorporate feedback to update style guidelines
Seniority
Mid-level, hands-on IC