We’re rolling out production-grade AI receptionists for small and mid-size companies in Egypt and I need an engineer who has already shipped a similar solution, not a proof-of-concept. The assistant must speak and understand Egyptian Arabic naturally in both voice calls and chat messages, handle the usual front-desk workloads (inbound calls and messages, lead qualification, appointment booking), and integrate smoothly with our clients’ Calendars, CRMs, and WhatsApp numbers. Robust fallback logic is essential: when confidence drops or a request sits outside the training data, the system should either clarify, escalate, or hand the thread to a human without losing context. Production reliability, call quality, and data security all matter more to me than flashy demos. A solid grasp of how Egyptian businesses actually schedule, track leads, and follow up will help you tune prompts, intents, and webhook flows accurately. To move forward, include in your proposal one real AI receptionist you have already deployed: what it replaced for the business, which stack or services you chose, and any link, video, or screenshots that prove it’s live. On delivery I’ll need: • A working voice-and-chat agent in Egyptian Arabic connected to Calendars, CRMs, and WhatsApp • A simple admin panel or documented APIs for updating FAQs and workflows • A short handover guide covering failover, human handoff, and how to add new flows If this first project goes well there’s a steady pipeline of similar builds. Show me your previous receptionist, and let’s start.