Samarkand delivery hub

Samarkand is Corpshore's second Uzbekistan delivery site and extends the company's rare-language and speech capacity beyond the capital. The owned hub draws on a historic Silk Road city with its own linguistic character, where Uzbek and Tajik are spoken side by side. That bilingual reality is useful for datasets that need genuine Tajik coverage rather than a Tashkent-accented approximation.

The site complements Tashkent by adding field-collection capability and additional recording capacity for speech corpora. Running two sites in one country gives programs redundancy and lets Corpshore balance load without moving data across borders. Teams here focus on transcription, annotation, and controlled speech capture.

As with every Corpshore hub, output moves through the three-tier QA cascade with in-region review, so dialect and script decisions stay with people who speak the language natively.

Samarkand's population is notably bilingual in Uzbek and Tajik, which is uncommon at scale and directly useful for Tajik speech and text datasets. Local teaching colleges and universities supply contributors comfortable with formal orthography in more than one language.

The city's contributor base rounds out Corpshore's Uzbek dialect coverage, since Samarkand Uzbek differs from the Tashkent and Fergana variants that models need to recognize.