Writing
Notes from building phone software for medical practices.
The arithmetic behind front desk costs, what HIPAA actually requires of a vendor, and what integrating with an EHR looks like from the inside. Written by the founder, not by a content team.
Latest
Which AI receptionists integrate with athenahealth? How to check any vendor yourself
A logo on a website is not an integration. Here is what a real one writes back to athenahealth, how to check a vendor on the athenahealth Marketplace in a few minutes, and what to watch on a demo.
Agni Patel, 8 min read
Medical practice phone statistics: missed calls, no shows, staffing and admin time, verified
Every figure here traces to MGMA, the AMA, a federal agency or a peer reviewed journal, with its population and year. The popular missed call numbers did not survive that test, and we explain why.
Agni Patel, 10 min read
AI receptionist vs IVR: should a medical practice replace its phone tree?
A phone tree sorts callers. An AI receptionist finishes what they called about. Most practices do not have to pick one, and some should keep the menu exactly as it is.
Agni Patel, 8 min read
Are AI receptionists safe for medical practices? What can go wrong, and the guardrails that prevent it
Five things can go wrong when software answers a practice's phones. Each one has a specific guardrail, and each guardrail can be tested before a single patient calls.
Agni Patel, 9 min read
Is an AI receptionist worth it for a medical practice? How to calculate the ROI
One formula, five inputs you can pull from your own phone system, a worked example with every assumption shown, and the cases where the answer is no.
Agni Patel, 9 min read
What does EHR write back mean for an AI receptionist, and why does it matter?
Read access lets a system look things up. Write back lets it finish the job. The difference decides whether an AI receptionist removes work or just moves it.
Agni Patel, 8 min read
Do patients need to know they are talking to an AI? Disclosure rules for medical practices
What California, Utah and the FCC actually require as of October 2026, what they do not, and a disclosure line you can use on every call.
Agni Patel, 9 min read
What is MedPhone AI? What it does, what it costs, what it integrates with, and who it is for
A plain product profile: the calls MedPhone AI completes, the ones it hands to staff, its integration status by EHR, published pricing, HIPAA posture and the practices it does not suit.
Agni Patel, 8 min read
AI receptionist for a medical office: what it picks up between 7:30am and close
Not the theory. The 8am line, the check in window, the refill calls after lunch and the 4:45 reschedule, and which of them an AI receptionist should take off your front desk.
Agni Patel, 9 min read
AI receptionist for a dental office: recall, hygiene columns, toothaches and the integration questions to ask
A dental phone line is not a medical phone line with teeth. The schedule runs on chair time, the money runs on recall, and the hardest call of the day is somebody in pain.
Agni Patel, 9 min read
Best AI receptionist for healthcare in 2026: a buyer's scorecard, not a ranking
Most lists of the best AI receptionists are written by somebody selling one. So is this one, so instead of a winner you get the seven things that actually separate vendors, and how to test each one yourself.
Agni Patel, 8 min read
AI voice agents in healthcare: what they are, real calls by department, and where they fail
A voice agent is software that holds a phone conversation and does something in a real system while it talks. In healthcare that is useful for a narrower set of calls than the marketing suggests, and the line is worth knowing.
Agni Patel, 8 min read
HIPAA compliant answering service: what the label can and cannot mean, human or AI
There is no official HIPAA certification for an answering service. There is a BAA, a set of safeguards, and six places a patient's information ends up after a call. Check all six.
Agni Patel, 8 min read
Medical receptionist duties: the full list, where the time goes, and what software should take
The job description says greet patients and answer phones. The real list is longer, and the phone sits underneath almost every line of it.
Agni Patel, 9 min read
What software do medical receptionists use? The front desk stack, layer by layer
Seven kinds of software, one of which matters far more than the rest. The test for every other tool is whether it writes back to that one.
Agni Patel, 9 min read
Medical office phone system: what it costs, the features that matter, and where AI fits
Most practices shop on price per seat and then live with the routing for seven years. Choose on queues, after hours rules and reporting instead.
Agni Patel, 9 min read
What is a nurse triage line? Who staffs it, how it works, and how it pairs with an AI front desk
Triage is clinical judgment over the phone. Routing is sorting calls by rule. Most phone safety problems at a practice start where the two get mixed up.
Agni Patel, 9 min read
Doctors office voicemail greeting scripts, plus the HIPAA safe messages to leave for patients
Ready to paste scripts for the main line, lunch, after hours, holidays and on hold, and the callback messages staff can leave without saying too much.
Agni Patel, 9 min read
AI scribe vs AI receptionist: two different jobs, and which to buy first
One works in the exam room and writes the note. The other works on the phone line and books the visit. They solve different problems, and most practices have one problem worse than the other.
Agni Patel, 9 min read
AI for medical practices in 2026: what works, what is early, and what to leave alone
Phones, notes, the inbox, billing, prior auth and clinical tools, each with an honest maturity read. Most of the value today is administrative, and that is fine.
Agni Patel, 10 min read
AI receptionist for mental health practices: intake, scheduling and crisis calls
The administrative calls at a therapy practice are ordinary front desk work. The crisis calls are not, and they never stay with the software. Get that rule right before anything else.
Agni Patel, 9 min read
Healthcare call center AI: what changes in central scheduling, and what to measure
For groups and health systems running a shared scheduling line. The routine calls leave the queue, the calls left for agents get harder, and the metrics need resetting to match.
Agni Patel, 9 min read
AI appointment scheduling for medical practices: what it books and what it should not
How an AI scheduling assistant books, reschedules and cancels by phone inside the EHR, where it should still hand the call to a person, and six test calls to make before you buy.
Agni Patel, 8 min read
How to handle prescription refill requests by phone without clogging the front desk
What a refill call needs to capture, where it should land, which ones go straight to a person, and what an AI phone agent can and cannot do with them.
Agni Patel, 7 min read
After hours answering for medical practices: what to cover, and who should answer what
Nights and weekends carry two different kinds of call. The urgent ones need a clinician. The routine ones need finishing, not a message for Monday.
Agni Patel, 7 min read
Bilingual AI receptionist for medical practices: what to test before you trust one
A second language on the phone line is an access question with legal obligations behind it. How to test a vendor properly, and what a phone agent should never be used for.
Agni Patel, 8 min read
How to reduce patient no shows at a medical practice
Reminders help, and the research says by how much. The lever most practices miss is making it easy to cancel or reschedule at the moment the patient decides not to come.
Agni Patel, 7 min read
Is an AI receptionist worth it for a small medical practice?
For a solo physician or a two provider clinic with one person on the desk, the answer turns on three numbers, and one of them is whether you are still taking new patients.
Agni Patel, 8 min read
Medical front desk burnout: what causes it and what to take off the desk first
The front desk is the job in a practice with the most interruptions and the least control over them. Pay matters, but the phone matters more than most practices think.
Agni Patel, 8 min read
AI receptionist for specialty practices: what changes from primary care
Orthopedics, dermatology, OB GYN, cardiology and ENT share half their phone calls with a family practice. The other half is where the configuration work lives.
Agni Patel, 8 min read
Insurance calls at a medical practice: what can be automated and what cannot
Some of the insurance calls a front desk takes are simple and some are not. Software vendors tend to blur that line. This post draws it.
Agni Patel, 7 min read
Virtual medical receptionist: remote staff, answering service or AI?
Three different products are sold under the same name. They differ less in price than in what is waiting for your front desk the next morning.
Agni Patel, 8 min read
How to reduce missed calls at a medical practice
Five fixes, four of which cost nothing and involve buying no software at all. Do those first.
Agni Patel, 9 min read
Which EHR systems work with AI phone agents?
A reference for the whole landscape, including the systems we do not support, plus the three integration depths that every vendor describes with the same word.
Agni Patel, 9 min read
What is a good containment rate for an AI phone agent?
The metric that decides whether a deployment worked, why launch always understates it, and the definition games vendors play with the word.
Agni Patel, 7 min read
How to deploy an AI phone agent at a medical practice
Five steps, a realistic six week timeline, and the four things that actually go wrong. None of them are technical.
Agni Patel, 10 min read
Is AI phone software HIPAA compliant?
It depends entirely on the vendor, there is no certificate anyone can show you, and here is how to tell the difference in one conversation.
Agni Patel, 8 min read
Can AI replace a medical receptionist?
Partially, and only the phone half. Here is the job broken into its actual tasks, with an honest mark against each one.
Agni Patel, 8 min read
How much does an AI receptionist cost for a medical practice?
The four pricing models vendors use, what sits outside the headline number, and the call volume below which none of it pays for itself.
Agni Patel, 8 min read
How does an AI phone agent work in a medical clinic?
The four things that happen between a patient dialing and an appointment appearing in the schedule, with the timings and the failure points.
Agni Patel, 9 min read
What is an AI receptionist for medical practices?
A straight answer to the definitional question, including the two things it is constantly confused with and the four jobs it still cannot do.
Agni Patel, 8 min read
How many calls does a medical practice actually miss?
The published figures run from 23 percent to 42 percent. Both get quoted as fact. Here is why they disagree and how to get your own number in an afternoon.
Agni Patel, 6 min read
AI receptionist or answering service: what actually changes
Both of them pick up at 9pm. Only one of them means your front desk has nothing waiting when they walk in at 8.
Agni Patel, 7 min read
Building on the athenahealth API: what an AI phone agent can actually do
What "integrates with athenahealth" really means when a vendor says it, from the side that had to build it.
Agni Patel, 10 min read
HIPAA compliance for AI voice agents: a buying checklist
There is no HIPAA badge and no HIPAA license. Any vendor can claim compliance. Here are the seven questions that separate the ones telling the truth.
Agni Patel, 8 min read
What a front desk hire actually costs a primary care practice
The salary line is about $35,000. The real number is closer to $52,000, and the gap is where most practice budgets quietly go wrong.
Agni Patel, 9 min read
How much is your practice losing to missed calls?
Four numbers you already have, and the arithmetic that turns them into a figure your P&L never shows you.
Agni Patel, 7 min read
Rather run the numbers than read about them?
Bring fifteen minutes and your call reports and we will do the arithmetic on your practice instead of a hypothetical one.
Book a 15 minute demo