Tashkent delivery hub

Tashkent is Corpshore's largest Central Asian delivery hub and the operational anchor for the company's rare-language and speech work. The owned facility runs full-time local teams rather than anonymous crowd labor, which lets projects keep the same annotators across long-running speech and preference-data programs. Uzbek, Russian, Karakalpak, and Tajik coverage is native and in-region, so pronunciation, dialect, and script decisions are made by people who use the languages daily.

The hub sits inside Corpshore's wider Uzbekistan footprint, which pairs Tashkent with a second delivery site in Samarkand. Teams here handle text-to-speech corpus design, transcription, and reinforcement-learning-from-human-feedback tasks that need consistent judgment over weeks. The city is Uzbekistan's administrative and technology center, which keeps recruiting, training, and QA supervision close to a deep pool of graduates.

Work is organized around Corpshore's three-tier quality cascade, so speech and annotation output is reviewed in-region before it reaches a client. That local review layer matters most for low-resource languages, where a distant reviewer cannot reliably catch dialect and orthography errors.

Uzbekistan has a young population with a large annual cohort of university and vocational graduates, many trilingual across Uzbek, Russian, and a growing base of English. That mix suits speech capture, transcription, and multilingual annotation where script switching between Latin and Cyrillic Uzbek is routine.

Because Uzbek is spoken across distinct regional variants, Corpshore recruits for dialect depth rather than a single standard accent, which is what TTS and ASR datasets need to generalize across real speakers.