We rely on three types of sources:
• Deep interviews with real people, conducted locally by experienced researchers. These cover values, behaviors, motivations, and choices. Responses are digitized and anonymized before training AI twins.
• Open data, including national statistics (e.g. stat.uz, UNDP, IMF), trend reports, media analysis, and public consumption data.
• Your own internal data, like CRM exports, website analytics, or survey results.