Voicebot for customer requests with CRM integration
I built a voicebot that takes customer requests over the phone and routes them to the responsible team.
- 500+calls per month
- over 50%of requests captured completely without manual rework
- 60+ hrsof working time saved per month
The situation
Phone requests need attention and follow-up. Employees have to balance their ongoing tasks with taking and routing these requests.
The specific task
Take customer requests over the phone and connect them with the responsible workflows in Casavi and the internal CRM.
My role
As AI project lead, I designed the voicebot, built it myself, migrated it and integrated it with Casavi and the internal CRM I developed.
Tools used
Rollout & usage
- 01 / PROTOTYPE
ElevenLabs + n8n
- 02 / MIGRATION
fonio.ai + Power Automate
- 03 / INTEGRATION
Casavi + self-built internal CRM
- 04 / DAILY WORK
In production use
Changes studied over six months.
The architectural approach: Connect voice interaction with downstream processing: the voicebot takes the request; the integration with Casavi and the internal CRM brings the information into the responsible workflows.
Try the case demoFrom conversation to case.
Routed automatically.
The following interactive sketch illustrates a possible flow from the prototype phase with ElevenLabs and n8n. Checks, branches and node names are examples. The current production solution uses fonio.ai and Power Automate; its internal configurations are not shown.
After the call, ElevenLabs hands over the transcript, phone number and time.
Illustrative example flow from the former prototype phase. The current production solution should be considered separately. No real AI calls, company data or internal configurations.
The difference in daily work.
More room for focused work
The benefit shows in daily work: phone requests reach the responsible team, while employees can concentrate better on their ongoing tasks.
Measurement scope: six months of daily work
The reported saving of 60+ working hours per month is based on a time comparison. Changes were studied over six months, focusing on uninterrupted work and the capture and routing of phone requests. Automatic capture means recording a request; subsequent handling is a separate process.
What it means in practice
Quality control & limits
Review questions for the example flow: Were the request and contact details captured correctly? Did the handover arrive? What happens with unclear information or an unreachable interface?
RAG knowledge system with Microsoft Copilot Studio
Let's
get in touch.
Looking for someone who connects AI applications with existing business processes? I look forward to talking about roles in AI consulting, automation and operational project management.
Get in touchThe button opens your email program.
mohamad@aboualnaser.com