Call center ai · Production

Retell AI automates 8,000+ monthly calls for Matic Insurance with 85–90% transfer success rate

The problem

Matic's call operations faced mounting inefficiency across 120,000 monthly calls: after-hours calls were dropped or mishandled by third-party vendors, agents spent 7–9 minutes per call on repetitive data collection before any consultative work, and scheduling delays caused missed appointments and lost high-intent leads.

First attempt

Third-party call center vendors handling after-hours calls delivered poor performance, with volume constraints causing missed calls, inconsistent messaging, and high-intent leads slipping through the cracks overnight.

Workflow diagram · grounded in source
1
Incoming call arrives
Trigger
An incoming call arrives after hours or as a scheduled appointment follow-up.
▾ source quote
“Matic launched an after hours AI phone agent to handle all incoming after-hours traffic”
2
AI collects contact and insurance info
Ai action
The AI phone agent collects basic contact and insurance info and schedules follow-up calls.
▾ source quote
“the AI phone agent collects basic contact and insurance info and schedules follow-up calls. This ensures that high-intent customers are never lost overnight and that Matic's representatives have all necessary context before following up.”
3
Appointment confirmation or rescheduling
Ai action
The AI voice agent calls the customer exactly on time, confirms availability, and reschedules if needed.
▾ source quote
“The AI voice agent calls the customer exactly on time, confirms they're still available, reschedules if needed”
4
Data intake and lead qualification
Ai action
The AI phone agent collects all 20–30 required data points and flags any disqualifying factors.
▾ source quote
“The AI phone agent collected all 20–30 required data points, flags any disqualifying factors, and hands off only eligible leads to licensed agents”
5
Transfer to licensed human agent
Routing
The call is transferred to a licensed human agent after AI qualification is complete.
▾ source quote
“transfers the call to a licensed human agent”
6
QA-driven improvement loop
Feedback loop
QA insights drive ongoing improvements to the AI voice agents.
▾ source quote
“There were five or six improvements we made just last week based on QA insights. This work is ongoing and essential.”
Reported outcome

Matic handled 8,000+ calls with AI in Q1 2025, achieved an 85–90% transfer success rate for appointment calls, automated ~50% of low-value tasks, maintained an NPS of 90 throughout the rollout, and saw 80% of customers complete AI-handled calls without requesting a human agent.

Reported metrics
Monthly call volume120,000
calls handled by AI in Q1 20258,000+
Transfer success rate for scheduled appointment calls85–90%
Call handling time reduction in data intake flows~3 minutes
Show all 11 reported metrics
monthly call volume120,000
calls handled by AI in Q1 20258,000+
transfer success rate for scheduled appointment calls85–90%
call handling time reduction in data intake flows~3 minutes
low-value tasks automated and reassigned~50%
NPS maintained throughout automation rollout90
customers completing AI calls without requesting a human80%
cost per policy and cost per transfersignificantly reduced
answer rate: AI vs. human agentshigher rate of answer rate using the bot
agent data-gathering time per call before automation7–9 minutes
call operations development focused on AI by early 202560%
Reported stack
Retell AITwilio
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.retellai.com/case-study/matic-insurance-ai-call-automation-case-study
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Matic handled 8,000+ calls with AI in Q1 2025, achieved an 85–90% transfer success rate for appointment calls, automated ~50% of low-value tasks, maintained an NPS of 90 throughout the rollout, and saw 80% of customer…

What tools did this team use?

Retell AI, Twilio.

What results were reported?

Monthly call volume: 120,000; calls handled by AI in Q1 2025: 8,000+; Transfer success rate for scheduled appointment calls: 85–90%; Call handling time reduction in data intake flows: ~3 minutes (source-reported, not independently verified).

What failed first in this deployment?

Third-party call center vendors handling after-hours calls delivered poor performance, with volume constraints causing missed calls, inconsistent messaging, and high-intent leads slipping through the cracks overnight.

How is this call center ai AI workflow structured?

Incoming call arrives → AI collects contact and insurance info → Appointment confirmation or rescheduling → Data intake and lead qualification → Transfer to licensed human agent → QA-driven improvement loop.

Related call center ai cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Call center ai cases →