Skip to content
Aldridge Dagos Get in touch

CASE 03

Healthcare Call Operations

A medical practice used to learn about missed calls and unhandled voicemails after the fact. Now it watches every call, queue, and staff member live.

Role
Call operations, queue design, and reporting
Client
A multi-site medical practice
Year
2026
Stack
Web app · Real-time · Telephony
Status
Live in production

01 The problem

A multi-site medical practice ran phone operations across many staff and queues with no live view of what was happening. A missed patient call is lost revenue and often a lost patient. The practice found out about misses the next day, long after the caller had given up and dialed someone else. Voicemails sat unhandled and nobody knew which ones still needed a return call. End-of-day numbers were guesswork. Nobody could say how many calls came in, how many got dropped, or who handled what. The team was flying blind on the one channel patients use most.

02 The system

A real-time call dashboard. Every call, every staff member, every queue, visible the moment it changes. The phone system pushes each call to the board as it happens, so the team watches the floor live instead of reading about it the next morning. A watchlist surfaces the calls nobody picked up and the voicemails still waiting, so a miss gets caught while there is still time to call the patient back. The watchlist also knows the difference between a real miss and an excused one: a call that rang someone already on the line or set to do not disturb does not count against them, so the list shows only what needs action. Every voicemail is tracked to done, returned or closed out with a logged attempt within one business day, instead of leaving a row stuck pending forever. Any call can be reviewed in place, with the recording playing inside the board in about a second and a half and every playback writing an audit entry before the first byte streams. An end-of-day report closes the day on its own. It sums the calls, the misses, and each person’s handle time without anyone pulling the numbers by hand.

03 How it holds up

The repo carries 1,301 tests across the call lifecycle, enrichment, reconciliation, and watchlist sort. Phone signals can arrive out of order, so the system reconciles questionable labels and durations against the authoritative phone log. Daily totals keep inbound queue traffic separate from outbound and after-hours activity, while unclassified calls remain named on the board. A watchdog checks the feed every five minutes and alerts on degradation and recovery, following the fail-visible standard for operational dashboards. Recording access uses rate limits and an audit log. Those controls support access review but do not, by themselves, claim HIPAA compliance.

04 The result

The practice sees its phones live instead of finding out after the fact. A missed call gets caught the same hour, while a staff member can still pick up the phone and win the patient back. End-of-day numbers stopped being guesswork and became a report the owner trusts. The system runs without me watching it and fixes its own records along the way.

05 Frequently asked questions

How do you catch a missed patient call before the patient is gone?

A watchlist on the live board surfaces every call nobody picked up and every voicemail still waiting. Instead of reading about a miss the next day, staff see it the same hour, while a team member can still pick up the phone and win the patient back.

How does a real-time call dashboard stay accurate when the phone system sends bad data?

The phone system fires signals fast and out of order, and the first one is often wrong. The dashboard treats the authoritative phone log as the source of truth and corrects its own record from it. Every call also lands in exactly one bucket of the daily totals, and anything unclassifiable is flagged by name instead of absorbed.

What happens if the dashboard’s data feed breaks?

A separate watchdog checks the feed every five minutes and sends an alert when it observes degradation, then confirms when it observes recovery. The board also flags totals it cannot reconcile and names the calls in question, reducing the risk that a feed gap looks like a quiet day.

How are patient call recordings protected?

Recording access sits behind a rate limit and an audit log. Access is capped in pace and leaves a record of who opened what, and the connection to the phone system runs server to server. That keeps patient calls protected while staff can still review the ones they need.

What goes into the end-of-day report?

The report closes the day on its own. It sums every call that came in, the ones that were missed, and each staff member’s handle time, then delivers itself each morning. Nobody pulls numbers by hand, and the totals come from the corrected call record, not guesswork.

· How it works

  1. 01

    A call comes in

    The phone system sends each call to the dashboard the moment it happens.

  2. 02

    The board updates live

    Every call, staff member, and queue shows its real state as it changes, not at end of day.

  3. 03

    Surface missed work

    A watchlist surfaces unanswered calls and voicemails still waiting for follow-up.

  4. 04

    The record self-corrects

    When a call looks mislabeled or a duration is wrong, the system fixes it from the authoritative record.

  5. 05

    The day closes itself

    An end-of-day report sums the calls, the misses, and each person's handle time on its own.

· Results

Tests
1,301
Status
Live in production
Built
Solo

1. 1,301 tests cover the call lifecycle, direction inference, enrichment, reconciliation, and the watchlist sort.

2. Server-to-server auth, updates driven from the authoritative phone record, and recording access behind a rate limit and an audit log.