Build a Custom AI Phone Receptionist & Order Intake System (Fixed-Price)

Project Overview: We are seeking an experienced AI voice developer to build a production-ready, inbound AI voice receptionist/agent on a per-project basis. The agent must handle live incoming phone calls with ultra-realistic, human-like conversation flow, collect structured details from callers, and push that data into an external database/workspace. ​Core Technical Requirements: ​Telephony & Engine Stack: Integrate a standard SIP/telephony provider (e.g., Twilio) with a low-latency voice framework (e.g., Vapi.ai, Retell AI, or LiveKit). ​Human Conversational Rhythm & Latency: ​Response latency must strictly be between 0.5s and 1.0s to avoid unnatural AI pauses. ​Fine-tune Voice Activity Detection (VAD) so the bot waits for the caller to finish speaking without cutting them off mid-sentence. ​Voice & Tone Configuration: ​Set up a clear, natural voice profile with an easy toggle to switch voice IDs (e.g., male/female options). ​Tone must be balanced, professional, and calm—never overly bubbly, monotone, or robotic. ​Dynamic Data Collection & Output Integration: ​The agent must follow prompt logic to listen to incoming caller requests, ask relevant clarifying questions to gather required details, and compile the final summary. ​Once a call finishes, the gathered information must automatically push to a designated web API (e.g., Google Workspace/Sheets or custom webhook) and trigger a notification. ​Scope & Project Delivery: ​Specific prompt instructions, business context, and data fields will be provided prior to contract start. ​Must deliver a fully working voice line for end-to-end testing, followed by code/account transfer. ​Budget & Proposal Requirements: This is a Fixed-Price contract. Please include in your proposal: ​Your fixed price bid to build and deploy this stack. ​The voice/telephony tools you recommend for sub-second latency. ​A link to an audio sample or video demo of a live voice bot you have previously configured.