AI Transcription Data Contributor

I’m assembling a small team to enrich our Bodo language AI models with fresh, high-quality transcription data. Your role is simple yet critical: listen to short Bodo audio snippets and produce precise, clean text transcriptions that follow our formatting guidelines.

You’ll be working exclusively with text data; no images or other media are involved. Accuracy and consistency matter far more than sheer speed, so I’m looking for someone who already has hands-on experience transcribing low-resource or regional languages. If you’ve handled dialect variations, speaker overlaps, or noisy recordings before, please mention that background—experience is the one detail I need to evaluate your fit.

The workflow is web-based, and you may use any keyboard tool or text editor you prefer, provided the final output is submitted in UTF-8 plain-text format. Regular feedback rounds will keep quality on track, and I’ll supply clear labeling rules up front.

Deliverables (per batch):
• Fully-transcribed Bodo text files, timestamp-aligned where required
• A brief notes file flagging uncertain words or audio issues

I release payment upon a quick spot-check that transcripts meet guideline accuracy and are free from spelling or alignment errors. If you enjoy meticulous language work and can commit at least a few hours each week, I’d love to hear about your relevant transcription experience and approximate weekly capacity.