DocsMD™Patient Revenue LibraryFree leak audit
DocsMD › Library › AI in the Front Office
AI in the Front Office

What does an AI receptionist do, and what does it cost?

What does an AI receptionist do, and what does it cost? Those are the two questions most practice owners ask first, and the answers are often blurred by demos. Put simply, an AI receptionist answers and routes calls. This guide explains what that covers, how pricing tends to be structured, the costs that don't show up on the quote, and how answering calls differs from recovering revenue.

Updated October 22, 2026 · 5 min read · DocsMD Research
Key takeaways
  • An AI receptionist is built to answer and route calls: greet, answer common questions, take messages, transfer, and sometimes offer times.
  • It works on inbound contact. Cancellations, no-shows and overdue patients don't call in, so they need outreach that starts on its own.
  • Total cost includes more than the plan: setup, scripting, integration, staff time on transfers, and how usage is metered.
  • Judge any front-office tool by the visits it puts back on the schedule, itemized, not by the calls it handles.

What an AI receptionist is built to do

An AI receptionist is software that answers your phone line, and sometimes web chat, using a voice or text model trained to hold a conversation. In a practice, the typical jobs are:

  • Answering every call, including overflow, lunch and after hours, so callers don't hit a busy signal or voicemail.
  • Answering routine questions: hours, location, parking, which insurance plans the office accepts, how to get forms.
  • Routing: sending billing calls to billing, clinical calls to a nurse line, and urgent calls to a person, following rules you set.
  • Taking messages and logging them for staff to return.
  • Offering appointment times, in some products, if it can see open times in your scheduling system.

Where its job ends

Every one of those jobs starts when someone calls. That's the design: an AI receptionist reacts to inbound contact. It does that job well when it's set up carefully, and it can take real pressure off a short-staffed front desk.

But a lot of lost revenue in a practice never rings the phone. A patient cancels by portal at 7:30 AM and a 10:00 slot sits empty. A patient doesn't show and nobody follows up. A patient who was due for a follow-up six months ago simply never hears from the office. None of these creates a call to answer. Recovering them takes software that notices something happened, or didn't, in the schedule and starts the outreach itself.

Our guide to AI receptionists vs. Patient Revenue Engines goes deeper into that difference. The short version: answering is about the conversation; recovery is about the visit.

How AI receptionist pricing tends to work

Prices vary widely and change often, so ask each vendor for a written quote based on your own call volume. What's useful is understanding the structures you'll see, because the structure decides what you'll actually pay:

  1. Usage-based. Charged per minute or per call. Simple to start, but your bill rises with call volume, which is highest in the weeks you're busiest.
  2. Tiered plans. A monthly fee covering a set number of minutes or calls, with overage charges beyond it. Ask what overage costs and how often practices like yours exceed the tier.
  3. Per location or per line. A fixed fee per phone number or office. Predictable, but check what's included.
  4. Setup and onboarding. Some vendors charge to build call flows, scripts and routing rules.
  5. Integrations. Connecting to your EHR or scheduling system to offer times may cost extra, or may only be available for certain systems.

The costs that don't show up on the quote

The plan price is rarely the whole cost. Before you compare quotes, add these up:

  • Staff time on transfers and messages. Every call the system routes to a person or turns into a message still needs a human. If most of your calls are clinical or billing questions, a large share comes back to the desk anyway.
  • Scripting and upkeep. Hours, providers, insurance lists and holiday closures change. Someone has to keep the system's answers current.
  • Errors and rework. A wrong answer about insurance or a booking in the wrong appointment type costs staff time and patient goodwill. Ask how the system handles questions it shouldn't answer.
  • Compliance work. Any vendor that hears or records patient calls handles protected health information and must sign a Business Associate Agreement with the practice before going live. Review what it records, transcribes and keeps.

A simple cost worksheet

The right way to judge cost is against what comes back. Here is illustrative math, not a quote from any vendor:

Say a practice misses 150 calls a month and an AI receptionist answers all of them. Of those, 60 are scheduling calls, and 30 end with a booked, kept visit that would otherwise have been lost. At an illustrative $140 per visit, that's $4,200 a month in recovered visits. Subtract the tool's monthly cost and the staff time still spent on transfers and messages, and you have its real return.

Now run the same sheet on the quiet leaks. Count last month's cancellations that were never refilled, no-shows that were never rebooked and patients more than three months overdue. For many practices, those numbers are larger than the missed-call number, and no phone tool touches them. Our cost of a missed call guide helps with the first half of the worksheet.

What actually recovers revenue

The front-office work that puts visits back on the schedule looks like this:

  1. Texting back missed callers within a minute with open times, so calls that got past everyone still turn into bookings. See missed-call text-back.
  2. Refilling cancellations by offering the freed time to patients who want an earlier visit, fast enough that someone can take it.
  3. Rebooking no-shows the same week, before they drift.
  4. Reactivating overdue patients who were due for a follow-up and never booked.
  5. Proving the result with an itemized list: patient, date, action and dollar value.
How DocsMD handles this

DocsMD recovers the visits, not just the calls

DocsMD is a Patient Revenue Engine, not a phone system: it texts back missed callers, refills cancellations, rebooks no-shows and brings back overdue patients, starting from your schedule export. Patients reply by numbered menu, and most replies never reach an AI model. Pricing is $699 per location per month, the Founder Rate for every practice that subscribes by December 31, 2026, with no per-message or per-user fees, and you can start with a free 30-day Test Drive, with nothing charged for 30 days from go-live.

Questions

Can an AI receptionist book appointments?

Some can offer times if they connect to your scheduling system. Check which systems are supported and who confirms the booking, so nothing lands in the wrong appointment type.

Is an AI receptionist HIPAA compliant?

It can be used in a compliant way if the vendor signs a Business Associate Agreement and handles call recordings and transcripts appropriately. Ask for the BAA in writing before going live.

How much does an AI receptionist cost?

It depends on the vendor and how pricing is metered: per minute, per call, tiered plans or per location, sometimes with setup and integration fees. Get a written quote based on your own call volume.

Does an AI receptionist fill cancelled appointments?

Usually only when a patient calls about one. Filling a cancellation means noticing the open time and reaching out to patients who want it, which needs outreach that starts on its own.

© 2026 DocsMD™ · Patient Revenue Engine for Healthcare · Practice smarter.docsmd.comLibraryFree leak auditPricingHIPAA BAAPrivacy