Case study / Patient access
A more connected
front desk
How an outpatient behavioral health practice worked with Zentric AI to expand phone coverage and improve the way patient requests are handled.
Explore the results ↓Reported deployment results
Conversations rated
“handled well”
First build compared with the latest build, using automated post-call assessment.
22 of 39 → 47 of 51 conversationsShare of inbound calls
reaching AiVA
Coverage expanded as the assistant entered fuller rotation in the practice’s phone workflow.
Early Results → Later ResultsBuilds
Refined through real conversations. Successive versions incorporated operational feedback from the practice’s coordinators.
Test calls excluded from quality resultsPatient requests don’t stop when the front desk is busy.
The practice needed a more consistent way to handle after-hours and overflow calls and get each request to the right team.
AiVA operated behind the existing phone menu, answering when staff were unavailable. The work focused on practical administrative tasks: capturing messages, sharing visit links, and routing requests for follow-up.
Better calls start
with listening.
Coordinators identified the friction. Zentric turned their feedback into concrete improvements.
Recognize returning callers.
Returning callers were matched to their record using their phone number, helping streamline the verification workflow.
Capture the message directly.
AiVA took messages during the call instead of sending the request back as another callback task.
Make the next visit easier to reach.
Callers could request their telehealth visit link by text.
Send requests to the right team.
Medical messages were routed to the caller’s provider; therapy messages went to the therapy desk.
Each version
moved the work
forward.
The share of evaluated conversations meeting the automated “handled well” criteria increased across all three builds.
These are version-level observations from real calls. They describe operational performance and do not establish clinical outcomes or a controlled comparison.
Automated “handled well” rate
Evaluated conversations: 39 / 389 / 51 respectively. Latest build: 47 of 51 conversations met the criteria.
Automation works
best with a path
to people.
Handling a call is only one part of the experience. Follow-through matters, too.
The deployment included an automated post-call review designed to flag unmet requests for staff follow-up.
The report also identified work still needed around reliable human transfers and verification when someone calls on another person’s behalf. Those limitations remain part of the story.
The operational lesson: build the handoff as carefully as the conversation.
Behind the numbers
How these results were measured
The source report excludes internal test calls. “Handled well” is an automated after-call assessment, described in the report as directional. The three versions recorded 22/39, 290/389, and 47/51 evaluated real conversations meeting those criteria. Different versions had different sample sizes and deployment conditions. These results are observational; no independent quality audit, clinical benefit, cost savings, or staff-time reduction is established here.
Client anonymity and reporting scope
The practice and assistant names, exact deployment dates, and phone-system vendor have been omitted. Results are drawn from a client deployment report. No patient records, recordings, or quotations are included.
Start with your workflow
What could a more connected front desk look like for you? Explore where AiVA could support your team’s patient call workflow.
Discuss your workflow ↗