Polish Transcript Annotation for AI

I’m expanding an AI-training initiative and need a native Polish speaker who can listen to short audio snippets of everyday, general-conversation Polish and annotate the accompanying transcripts inside our web-based labeling tool. For every segment you will compare what the system thinks was said with what a fluent human would actually write, marking errors and supplying corrections.

Your focus will be on:
• Grammar accuracy
• Clarity of message
• Natural phrasing

Each transcript you return should read as if it were typed by an articulate native speaker, free of awkward word order or machine-translation mistakes. No specialised jargon or technical terminology is involved—the recordings revolve around ordinary topics such as greetings, small talk, directions and casual opinions.

I provide detailed guidelines, quick video walkthroughs of the interface, and a Slack channel for questions. Previous experience with linguistic annotation, ASR post-editing, or tools like ELAN, Transcriber or any custom tagging platform is welcome but not a prerequisite; patience, sharp attention to detail and flawless written Polish matter most.

Turnaround is flexible: batches are assigned daily, and you can work whenever suits you as long as deadlines are met. I review the first few files together with you so we align on style; after that, a spot-check system ensures consistency. If you enjoy language work and want to help push Polish NLP forward, I’d love to add you to the team.