Call center ai · Production

Medical Data Systems scales $280,000 in monthly collections with Retell AI voice agents

The problem

Inbound call volume at MDS exceeded 3,000 calls per day, with many going unanswered or to voicemail. High agent turnover, expensive training, and the inherent difficulty of scaling phone support without adding headcount made voice the hardest and costliest part of the operation.

Workflow diagram · grounded in source
1
Inbound patient call received
trigger
“these agents now form the first layer of MDS's contact center”
2
AI handles patient inquiry
ai_action
“One agent answers inbound patient calls, helping callers understand their balances, verify insurance payments, and complete transactions”
3
Route to live agent or complete
routing
“only about 30 percent of those calls needing to be transferred to a live agent. That means the AI completes roughly 70 percent of conversations end-to-end on its own”
4
Outbound secondary collections
ai_action
“Another agent makes outbound calls for secondary collections, reaching patients with lower-balance accounts to resolve payments quickly”
5
Insurance IVR navigation
ai_action
“A third agent calls insurance companies, navigates their IVR systems, waits on hold, and transfers live representatives to MDS staff when needed”
Reported outcome

Retell voice agents now handle 100 percent of MDS's inbound calls, completing roughly 70 percent end-to-end without human transfer.
Across inbound and outbound workflows the AI handles approximately 30,000 calls per month and collects an average of $280,000 per month, while call abandonment has not increased.

Reported metrics
peak inbound call volume (pre-AI)3,000 calls per day
inbound calls handled by AI100 percent
Calls transferred to live agentabout 30 percent
calls completed end-to-end by AIroughly 70 percent
Show all 7 reported metrics
peak inbound call volume (pre-AI)3,000 calls per day
inbound calls handled by AI100 percent
calls transferred to live agentabout 30 percent
calls completed end-to-end by AIroughly 70 percent
calls handled per monthapproximately 30,000
monthly collections$280,000 per month
call abandonment ratehas not increased
Reported stack
RetellIVR
Source
https://www.retellai.com/case-study/how-medical-data-systems-scales-280-000-in-monthly-collections-with-ai-voice-agents-on-retell
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Retell voice agents now handle 100 percent of MDS's inbound calls, completing roughly 70 percent end-to-end without human transfer.

What tools did this team use?

Retell, IVR.

What results were reported?

peak inbound call volume (pre-AI): 3,000 calls per day; inbound calls handled by AI: 100 percent; Calls transferred to live agent: about 30 percent; calls completed end-to-end by AI: roughly 70 percent (source-reported, not independently verified).

How is this call center ai AI workflow structured?

Inbound patient call received → AI handles patient inquiry → Route to live agent or complete → Outbound secondary collections → Insurance IVR navigation.

Related call center ai cases