Freelance Audio Contributor – US English

Perle Systems; United States; Remote; USD 25,000 - 30,000; Full time

Job summary

Perle Systems is seeking native US English-speaking university students to join its AI audio training pipeline as Freelance Audio Contributors. You will record conversational audio from provided scripts to feed training loops of next-gen AI systems.

Remote, project-based role built for students with flexible hours. You’ll work with clear audio specs, and each audio file will be 10 seconds to 1 minute. Compensation is $20 per hour of recorded audio, with fully remote work from anywhere in the US.

Qualifications

• Native US English speaker with a standard American accent.
• Access to a quiet recording environment and a reliable headset/microphone.
• Strong attention to detail and ability to follow written guidelines.
• Ability to work independently and meet batch delivery deadlines.
• Basic computer literacy (file management, audio uploads, spreadsheet tools).
• Prior experience in transcription, voice work, or data annotation is a plus but not required.

Responsibilities

• Record natural two-speaker conversations from provided scripts and prompts designed to train AI speech models.
• Keep dialogue sounding natural and spontaneous to help AI learn authentic human conversation patterns.
• Ensure recordings meet technical specs: minimum 16kHz sample rate, clear audio, minimal background noise.
• Each file must be between 10 seconds and 1 minute in length.

Skills

US English

Attention to detail

Independent work

Computer literacy

Transcription/voice work (nice-to-have

Education

University student

Tools

Spreadsheet tools

Job description

About Perle

Perle is a data sourcing company dedicated to building high-quality, human-generated datasets that power the next Perle AI builds the foundational training data that powers the world's most advanced AI models. We partner with leading AI companies to source high-quality, human-generated datasets that make AI smarter, more natural, and more capable. Our contributor network spans the globe, and our data directly shapes how AI understands and generates human language.

About the Role

• We're looking for native US English speaking university students to join our AI audio training pipeline as Freelance Audio Contributors. This is your chance to be on the front lines of how AI learns to understand, process, and generate human speech. You'll record conversational audio from provided scripts, feeding directly into the training loops of next-gen AI systems.
• Remote. Flexible. Project-based. Built for students.

What You'll Do

AI Voice Data Generation

• Record natural two-speaker conversations from provided scripts and prompts designed to train AI speech models
• Keep dialogue sounding natural and spontaneous to help AI learn authentic human conversation patterns
• Ensure recordings meet technical specs: minimum 16kHz sample rate, clear audio, minimal background noise
• Each file must be between 10 seconds and 1 minute in length

What You Need

• Native US English speaker with a standard American accent
• Access to a quiet recording environment and a reliable headset/microphone
• Strong attention to detail and ability to follow written guidelines
• Ability to work independently and meet batch delivery deadlines
• Basic computer literacy (file management, audio uploads, spreadsheet tools)
• Prior experience in transcription, voice work, or data annotation is a plus but not required

Why This Matters

Every major AI breakthrough starts with high-quality training data. The models behind voice assistants, AI-generated speech, and conversational AI all depend on real human voices and real human judgment. This isn't just a side gig. You're directly contributing to how AI evolves.

What's in it for You

• $20 per 60 minutes of recorded audio
• Fully remote, work from anywhere quiet
• Flexible hours that fit around your class schedule
• Hands-on experience inside a real AI data pipeline
• Understand how AI models are trained from the ground up

Important Notes

• All audio data must be original and legally sourced
• Submitted data must not include any personally identifiable information (PII) of third parties without explicit consent
• Perle retains all intellectual property rights to collected data and associated deliverables