This project involves segmentation followed by verbatim transcription of audio clips in the target language.
Starting date: ASAP, this week
Project duration: 6-8 weeks
No. of resources: 20 approved annotators
Languages:
Kuki-Chin-Mizo (Mizo)
Annotators will:
Segment audio into “Speech,” “Babble,” “Music,” or “Noise” segments with precise timestamps
Convert segmented speech into written text following strict formatting conventions
Adhere to comprehensive segmentation and transcription guidelines
Maintain ≥95% accuracy
Work exclusively on the client’s designated annotation platform
Purpose of the Project:
To create high‑quality training data for speech technology systems by accurately segmenting and transcribing spoken content into correctly formatted written transcripts.
Essential Qualifications:
Native‑level proficiency in the target language
Strong attention to detail and commitment to accuracy
Reliable internet connection and headset for audio playback
Ability to follow the guidelines consistently
Preferred Qualifications:
Prior experience in audio segmentation, transcription, data labeling, or linguistic work
Familiarity with technical annotation platforms
Understanding of linguistic concepts
Additional Information
Data security and confidentiality protocols must be followed.
Initial certification is required via OneForma 2.0. (There are two Certifications: one for segmentation and the other for transcription, they’ve been included in the job requests)
Long‑term collaboration is possible for qualified annotators.
Rate per hour applies. TPT (Time Per Task) will be clarified later on.
