I need a full-featured AI solution that answers, places and chats with customers as naturally as a seasoned support agent. The system must pick up inbound calls, initiate outbound follow-ups when certain triggers are met, and switch seamlessly to live chat when visitors use our web widget. Beyond basic FAQs, it should diagnose and resolve advanced issues—think multi-step troubleshooting, order amendments, and policy clarifications—without escalating unless absolutely necessary. Ideal build uses real-time speech recognition and synthesis (Twilio Voice, Amazon Connect or comparable), a robust NLU engine such as Dialogflow CX or GPT-4, and an orchestration layer that lets me map complex conversation flows, integrate CRM data, and insert business-specific logic written in Python or Node. I’ll provide sample transcripts, escalation rules and API documentation for our ticketing platform. Deliverables • Deployed voice and chat endpoints connected to my phone numbers and website • Custom intent/slot models trained on my dataset, capable of advanced problem solving • CRM/API integration for lookup, ticket creation and automated callbacks • Dashboard or logs that let me review full conversation histories and tweak flows • Short hand-over guide plus one live test session to confirm everything works end-to-end Acceptance criteria: average user satisfaction ≥85 % in the pilot week, first-call resolution ≥70 % on the provided issue set, and zero downtime during the agreed live demo. Let me know your preferred tech stack, past projects that prove you can reach these metrics, and an estimated timeline to go from kickoff to production.