Uzbek Speech Data Collector
Corpshore AI hires Uzbek Speech Data Collectors onsite in Tashkent, Uzbekistan on a gig basis to record and transcribe Uzbek, Russian, and Karakalpak speech for AI voice-model training. No experience is required; paid onboarding and studio equipment are provided, and pay is issued per approved recording batch.
Corpshore AI's Tashkent hub is the company's largest Central Asia delivery center and its center of gravity for rare-language speech data. As an Uzbek Speech Data Collector, you'll record scripted and spontaneous speech in our sound-treated studio and transcribe existing audio, building the datasets that train text-to-speech (TTS) and automatic speech recognition (ASR) systems in Uzbek, Russian, and Karakalpak. Recent Corpshore AI speech work in this hub cut word error rate (WER) on a multilingual TTS/ASR program from 11.2% to 4.3% — the kind of measurable outcome this role directly contributes to.
Key responsibilities
- Record scripted prompts and free-form speech samples in Uzbek, Russian, and/or Karakalpak in a studio setting
- Transcribe raw audio to text with precise timestamps, disfluency markers, and speaker tags
- Follow phonetic and prosody guidelines so recordings are usable for TTS voice model training
- Flag audio quality issues (background noise, mic clipping, mispronunciation) before batch submission
- Hit weekly recording/transcription quotas tracked in Jwuma without sacrificing accuracy
What we look for
- Native or near-native fluency in Uzbek; conversational Russian or Karakalpak a plus
- Clear, neutral speaking voice with no persistent illness-related vocal impairment
- Comfortable being recorded and reading scripted content aloud for extended sessions
- Available to work onsite at the Tashkent studio on a flexible gig schedule
- Basic computer literacy for logging sessions in Jwuma
Nice to have
- Prior voiceover, radio, call-center, or language-teaching experience
- Familiarity with a second Central Asian dialect (Tajik, Karakalpak)
- Experience with transcription or subtitling tools
Compensation and benefits
- Gig-rate pay per approved recording/transcription batch, paid weekly
- Free onboarding and phonetic-script training before paid work starts
- Access to Corpshore AI's Tashkent studio equipment — no personal gear required
- Path to contract or permanent QA/team-lead roles for top performers
Common questions
How do I get a job recording Uzbek speech data for AI in Tashkent?
Apply through Corpshore AI's careers page, complete a short paid-style voice and reading assessment, and if selected you'll be onboarded onsite at the Tashkent studio with no prior experience required.
Do I need my own recording equipment for this role?
No. All recording happens in Corpshore AI's sound-treated Tashkent studio using company-owned equipment.
Is this a remote job or an onsite job?
This is an onsite, gig-based role based at Corpshore AI's Tashkent, Uzbekistan hub.
Similar roles
Apply once, get matched to what fits
One application covers every role you qualify for. A person reads it.
Apply now