Front-desk support at $0.08 a call
Phone calls → transcribe, understand, route → Dynamics 365 + ticketing
An AI voice agent - ElevenLabs voice over Twilio, orchestrated in n8n with a custom MCP server - answers calls, captures intent, books consultations, handles tier-1 support, and writes outcomes straight into Dynamics 365 with transcripts and outcome tags, for roughly $0.08 a call. It's the same approach as everything else: extract structured data from one channel, validate it, and deliver it cleanly to another.
Results at a glance
- 24/7 automated booking and tier-1 support
- Met the ≥95% task-completion / containment target
- Real-time CRM logging with transcripts and outcome tags
- ~$0.08 per call (typically 60–90 seconds, usage-based)
- Demo-ready in 2 weeks, full rollout in 6
The problem
Human support staff were stretched managing CRM updates, scheduling, ticketing, and communication across email, phone, and chat. Repetitive work drained resources, slowed response times, and undermined the company's modern brand positioning.
What was built
Voice
ElevenLabs voice delivered over Twilio, so callers get a natural, responsive agent rather than a phone tree.
Orchestration
n8n coordinates the microservices through a custom Model Context Protocol (MCP) server that gives the agent its tools.
Integrations
Wired into Microsoft Teams, Dynamics 365 CRM, Outlook calendar, and custom AI tools, so bookings and outcomes land where the team already works.
Reliability
Tuned over roughly 3-day cycles to reach a ≥95% containment target, with Loom walkthroughs, documentation, and team training at handoff.
How the approach worked
A phone call is just another channel to extract structured data from: capture the caller's intent, validate it, and deliver it cleanly into the CRM. Get the containment rate high enough and the economics change completely.
The results
The agent launched on schedule. Basic support and scheduling workload dropped significantly, faster bookings and response times lifted customer satisfaction, and the team moved from repetitive tasks to higher-value case management.
High costs, error-prone integrations, slow bookings, hard to scale.
Major labor savings, seamless cross-system flow, fast self-serve, ready for outbound expansion.
Noah was fast, communicative, and cracked every integration, some of which we worried weren't even possible.
Tools & technologies
Have a process like this slowing your business down? Tell me your stack and goals and I'll send back a build plan.
Take the stack assessment →