AI Inbound Call Automation Platform

I’m building a sales-and-service automation platform and the first milestone is a robust AI inbound call module. The system must pick up live calls, recognise intent, and hold natural, contextual conversations that solve customer-support issues and answer general enquiries without handing off to a human unless confidence drops. Think of the experience people get with a well-trained Dialogflow CX or GPT-4-powered voice bot, only fully integrated into my sales stack.

Core requirements
• Voice interface that can greet the caller, authenticate if needed, and maintain context across multi-turn dialogue.
• Ability to fetch or write data to our CRM so answers are personalised (order status, account info, FAQs, etc.).
• Real-time sentiment and confidence scoring with graceful human escalation when necessary.
• Call recording, transcripts, and basic analytics available in a dashboard.

Preferred tools & integrations
I’m already licensed for Twilio Voice, HubSpot, and OpenAI, so leaning on those APIs will speed us up, but I’m open to comparable alternatives if you can show clear advantages.

Expansion roadmap
Once the inbound flow is stable we’ll move to outbound dialers, appointment scheduling, and lead-gen workflows, so clean, modular code and thorough documentation are essential.

Deliverables for this phase
1. Deployed voice bot accessible from a dedicated phone number.
2. Source code with install/readme instructions.
3. Post-deployment test report covering at least 20 live calls across key scenarios (support, general inquiry, fallback).

If you’ve shipped conversational voice bots that feel human and plug into CRMs, let’s talk about timelines and the best way to stage this build.