Commercially licensable datasets collected across 18+ countries and 35+ languages. Browse ready-made catalogue items or commission a dataset built to your exact specification.
Corpshore AI operates data collection hubs across four continents. In Latin America, our teams are based in Santo Domingo (Dominican Republic), Mexico City and Bogota, covering 17 Spanish dialect variants and Haitian Creole. In Eastern Europe, our Warsaw and Istanbul hubs deliver Polish, Ukrainian, Turkish and Kurdish datasets. Across West Africa, our Accra hub covers Twi, Ewe, Ga, Fante, Hausa, Yoruba, Igbo and Ghanaian English. Our Nairobi and Kampala hubs serve East African languages including Swahili, Luganda, Acholi, Kikuyu, Luo and Kinyarwanda. In Southeast Asia, our Manila and Ho Chi Minh City operations produce Filipino, Cebuano, Ilocano, Vietnamese and English datasets. And in Central Asia, our Samarkand hub is the only commercially active Uzbek, Tajik and Kyrgyz data collection operation in the region.
Don't see what you need? We build it. Describe your requirements and our data operations team will scope, collect and deliver a dataset built to your exact specification with full commercial licensing and consent documentation.
Corpshore AI offers a catalogue of commercially licensable AI training datasets collected across 18+ countries and 35+ languages. Datasets cover egocentric video, multilingual speech, NLP, human motion, image annotation and synthetic data. Browse the catalogue at data.corpshore.ai or contact info@corpshore.ai to commission a custom dataset.
Egocentric video is footage captured from a head-mounted or chest-mounted camera that shows the world from the perspective of the person performing a task. AI labs use it to train embodied AI, humanoid robot models and vision-language-action (VLA) models that need to understand and replicate human actions in real-world environments.
Yes. Corpshore AI collects data in Twi, Ewe, Ga, Fante, Hausa, Yoruba, Igbo, Swahili, Luganda, Acholi, Kikuyu, Luo, Luhya, Kinyarwanda, Dagbani, Amharic and Oromo through its Ghana, Kenya and Uganda delivery hubs. Both speech and text datasets are available, commissioned to your specification.
Corpshore AI offers speech data in 35+ languages including Spanish dialects across Latin America and the Caribbean, English variants from North America, West Africa and Southeast Asia, French (African and European variants), Polish, Turkish, Uzbek, Tajik, Kyrgyz, Vietnamese, Filipino (Tagalog, Cebuano, Ilocano), Swahili, Luganda, Twi, Hausa, Yoruba, Igbo, Arabic dialects, Hindi, Bengali, Mandarin Chinese, Japanese, Korean and more.
Corpshore AI datasets are available under non-exclusive perpetual and non-exclusive annual licenses. Exclusive licensing is available on request. All datasets are collected with full commercial rights assignment from contributors and comply with GDPR and applicable local data protection laws. IRB-ready consent frameworks are available for sensitive domains such as healthcare and children's data.
Yes. Corpshore AI commissions datasets to spec through its Jwuma contributor network spanning 18+ countries. Submit a brief via the commission form on this page or email info@corpshore.ai. The data operations team will scope, collect and deliver with full licensing documentation.
RLHF stands for Reinforcement Learning from Human Feedback. It involves collecting human preference judgements on AI model outputs to train reward models that improve AI alignment. Corpshore AI offers RLHF preference pair datasets in English, Spanish, French, Polish, Turkish, Vietnamese and Filipino, with domain-specific variants for legal, healthcare, finance and customer service.
Corpshore AI offers datasets at startup, growth and enterprise tiers. Minimum order sizes vary by data type. Contact info@corpshore.ai or use the quote request form on the product page to discuss your volume requirements.
Yes. Corpshore AI produces sentence-aligned bilingual corpora for machine translation training across English-Spanish, English-French, English-Polish, English-Turkish, English-Swahili, English-Vietnamese, English-Tagalog, English-Uzbek, Swahili-Luganda, Twi-Ewe, Twi-Hausa and other language pairs. All pairs are human-translated by native speakers.
Delivery timelines depend on dataset size, complexity and geography. Small commissioned datasets (under 100 hours or 50,000 items) typically take 2 to 4 weeks from brief to delivery. Larger programs are scoped individually. Contact info@corpshore.ai to discuss your timeline.
Corpshore AI collects data through delivery hubs in the Dominican Republic, Mexico, Colombia, Poland, Turkey, Ghana, Kenya, Uganda, Philippines, Vietnam and Uzbekistan (Samarkand), as well as through extended contributor networks across Latin America, West Africa, East Africa, Southeast Asia and Central Asia — 18+ countries in total.
Yes. All Corpshore AI datasets are collected under signed informed consent frameworks that meet GDPR requirements for EU and EEA buyers. Data is collected, stored and transferred in compliance with applicable local data protection law in each collection country. IRB-ready consent documentation is available for sensitive data categories.