I have already built an ElevenLabs-powered voice agent to answer every incoming call to the Ausrealty office. What I still need is a specialist who can make the back-end data connections work flawlessly and verify that call-routing behaves exactly as intended. Here’s the current situation: • The agent exists in ElevenLabs with its base prompt and voice. • It must draw live property listings, rental details, and agent/contact profiles from three places—our public website database, an external listings API, and our internal CRM. • Once the data is flowing, the agent must recognise the caller’s intent and transfer them immediately: deposit inquiries to Accounts, rental inquiries to Property Management, and sales inquiries to the correct suburb agent. Where I’m stuck is the data side. The website refuses to release listing data when scraped, so the agent cannot answer basic questions. Your first task is to diagnose and resolve that blockage (adjust scraping logic, use the site’s underlying API, or propose another reliable method). When the data feeds are stable, I then need you to wire them into the agent’s NLU layer, map the call flows, and run end-to-end testing to prove that: 1. Property, rental, and agent information is retrieved in real time from the website database, external API, and CRM. 2. The voice agent correctly identifies deposit, rental, and sales intents and pushes the call to the right SIP/phone endpoint without delay. 3. Handover logging shows the contact, time stamp, and intent for every transfer. Please outline the tools or frameworks you prefer for web scraping, API integration, and IVR/SIP routing, along with a timeline. Once confirmed, I’ll provide API keys, CRM access, and the current ElevenLabs configuration so you can dive straight in.