I need a Voice AI agent built from the ground up to handle customer support across three channels—phone systems, web applications, and mobile apps. Its core responsibilities are: • Instantly answering frequent questions with accurate, conversational responses • Detecting intent and, when necessary, routing callers or users to a live agent without losing context Here’s what I expect from the engagement: 1. A working voice interface for each channel, leveraging proven services such as Twilio, WebRTC, or equivalent for telephony, and Web-/mobile-friendly speech recognition APIs. 2. Natural-language understanding in English, tuned to my support knowledge base and FAQs, with fallback handling for unknown queries. 3. An admin dashboard or clear logging endpoints so I can review transcripts, see hand-off events, and refine intents. 4. Simple integration hooks (REST or GraphQL) that my existing backend can call to update answers or trigger call routing rules. I’ll provide the current FAQ list and sample call flows. You bring the voice AI expertise—dialogue design, intent training, speech-to-text, text-to-speech, and smooth live-agent transfer logic—packaged as a functional prototype ready for internal testing.