Case study · Healthcare provider

Patient support beyond office hours.

A healthcare provider's patients were reaching out by phone, chat and the portal. The answers were spread across four systems that didn't talk to each other. We connected them, put AI to work in five teams and kept every tool the staff already used.

Case file Patient experience
Client
Healthcare provider, name withheld
Scope
Scheduling, referrals, billing and patient support
Systems
EHR, ticketing, scheduling and an internal admin tool
Teams using it
5
Built
5 AI capabilities
Existing tools
All kept

Built around existing tools

What changed

What changed for patients and staff.

  • Referrals

    Patients hear back sooner about referrals

    Patients get a reply while they're still looking to book, before they've spent too long waiting in the referral queue.

  • Routine questions

    Simple requests answer themselves

    Appointment status, refills and rescheduling get handled on the spot. Harder cases go to a person with the full history attached.

  • Coverage

    Help after 5 p.m.

    Patients get answers in the evenings and on weekends, not just between 8 a.m. and 5 p.m.

  • Existing tools

    Staff kept the tools they knew

    The new workflows use the tools staff already knew, so nobody had to learn a new system.

The starting point

One patient, four disconnected systems.

Much of a patient's experience happens outside the exam room: booking or rescheduling a visit, following up on a referral or asking about a bill.

The details were spread across four disconnected tools. Before replying, staff had to check all four and put the patient's history together themselves.

We connected the tools so patients can get one clear answer even when the information comes from several systems.

System map 4 connected systems
  • EHR Records and visit history
  • Ticketing Messages and requests
  • Scheduling Appointments and changes
  • Internal admin Billing and back office

AirCaptain reads · routes · drafts · follows up

One answer for the patient phone · chat · portal

What we built

Five AI capabilities, built into the daily work.

Each capability works with the provider's existing tools, and we measure the results for each one.

  1. Triage and routing

    For every inbound message

    Every message gets read, tagged by what the patient needs, scored for urgency and sent to the right queue. Urgent cases move to the front instead of waiting their turn.

  2. Care team copilot

    For staff

    When someone opens a case, the patient's history from all four systems is already on one screen, with a drafted reply and a suggested next step. Staff don't have to pull the history together themselves.

  3. Patient-facing automation

    For patients

    Voice, chat and automatic replies for requests that never needed a person. Appointment status, booking and rescheduling, cancellations, referral status and refill status.

  4. Back-office follow-through

    For operations

    Once a case is resolved, the follow-up work that normally spans several tools runs on its own. Staff step in when something needs a closer look.

  5. Trends in patient sentiment

    For leadership

    The system analyzes every conversation. Leaders can spot trends as they emerge, get alerts when volume spikes and see how patient sentiment changes from week to week.

Teams using it

Five teams went live, each in the tools it already used.

  • Team 01

    Scheduling

    Booking, rescheduling and confirmations.

  • Team 02

    Referrals

    Inbound and outbound referrals, start to finish.

  • Team 03

    Billing and revenue cycle

    Claims, appeals and collections.

  • Team 04

    Medical assistants

    A copilot, plus prescription and prior authorization workflows.

  • Team 05

    Patient support

    Messages, questions and requests from every channel patients use.

Your operation

Recognize any of this?

Book a call directly, or send us a message and we'll set one up. On a 30-minute call we'll show you where AI fits in your stack and where it doesn't. If it's a fit, you'll see a working prototype on your real workflows in the first week.

Next: the retail case study