Aloqa is an AI receptionist for small businesses in Uzbekistan and Central Asia — cafes, shops, salons — that have no CRM and no support team. Customers write in Telegram, often as voice notes, and leave reviews on maps; owners miss complaints and lose customers. Aloqa collects incoming messages, transcribes voice notes, classifies intent, urgency and risk, drafts polite replies in Russian or Uzbek, and creates actionable tasks. The owner always approves every reply — human-in-the-loop, no biometrics, no emotion detection. Built entirely in native.builder, on a mobile phone. The AI layer runs live on Fireworks AI via Edge Functions, with a deterministic rule-based fallback, so the product never fails. Security audit completed: JWT verification, restricted CORS, input validation, legacy code removed. Speechmatics, AI/ML API, Featherless AI and Bright Data are wired as integration-ready stubs. Impact: 2–3 hours of daily message handling reduced to 20–30 minutes; fewer lost customers; consistent brand voice in two languages.
Category tags: