The message that lands at nine at night, and no one answers
A patient notices their near vision is blurry, searches for an "ophthalmologist" on their phone, and sends a WhatsApp message: "Do you do cataract surgery? How much is the evaluation?" It's nine at night. Your receptionist left at six. The message sits on read until mid-morning tomorrow, and by then the patient has already messaged two other clinics and booked with whoever answered first.
This happens every day in an ophthalmology practice. Demand isn't the problem. What's missing is someone to answer at the exact moment the patient decides to act, which rarely lines up with your office hours. Hiring a second receptionist to cover evenings, nights, and weekends is expensive and impractical. The real alternative is an AI agent that runs your WhatsApp in your practice's voice, around the clock.
What the agent handles from the very first night
The agent isn't a button-menu chatbot. It converses. When a message comes in, it replies the way your best receptionist would, because it knows your practice:
- It answers FAQs: evaluation prices, which procedures you offer (cataracts, LASIK, retina, glaucoma), location, which insurers you accept.
- It tells a real emergency ("I'm suddenly seeing floaters and flashes of light") from a routine inquiry, which is critical in ophthalmology.
- It books the appointment on your calendar without a person stepping in.
- It sends the evaluation payment link when relevant.
- It logs every conversation in your CRM, so you know who asked what.
The patient sees a clinic that replies instantly. You didn't hire anyone.
The real cost of "we'll answer tomorrow"
Let's run the numbers. A mid-sized ophthalmology practice can receive 15 to 30 new WhatsApp messages a day. A meaningful share arrive outside office hours: nights, Saturdays, Sundays. If just three of those messages a week go cold because no one answered in time, and the average evaluation plus the procedure that may follow are worth what they're worth in ophthalmology, the monthly lost revenue is uncomfortable to look at.
The problem isn't that your receptionist is slow. It's that she's human: she has set hours, she works the front desk, she answers the phone, and she can't be in three places at once. The agent covers exactly the gap she can't: the late night, the Sunday, the minutes she's with a patient in the room.
Human receptionist vs. AI agent on WhatsApp
| Aspect | Receptionist (set hours) | AI agent on WhatsApp |
|---|---|---|
| Response hours | 9 to 6, Mon-Fri | 24/7, weekends included |
| Response time | Minutes to hours | Seconds |
| Simultaneous messages | One at a time | Unlimited at once |
| Eye-emergency triage | Depends on her judgment | Consistent rules you defined |
| Record-keeping | In her memory or on paper | In the CRM, with full history |
| Cost | Recurring monthly salary | One-time build |
The agent doesn't replace your receptionist; it lifts the 30 repetitive messages a day off her plate so she can focus on the patient standing at the front desk.
Why ophthalmology needs this more than other specialties
There's a clinical reason on top of the commercial one. In ophthalmology, time matters. A retinal detachment, acute glaucoma, a chemical burn: each has a window in which the right response changes the outcome. A patient who writes "I suddenly see a black curtain over one eye" at eleven at night cannot sit on read until tomorrow.
The agent is built to recognize that alarm language and act on the rules you set: tell the patient to go to the emergency room immediately and, in parallel, escalate the conversation to a person on your team with full context. It doesn't improvise diagnoses or give medical advice it has no business giving. It knows how to tell urgent from routine and, above all, it knows when to hand the conversation to a human.
That combination — instant response for the routine, judgment to escalate the delicate — is what makes an agent safe in a specialty where an unanswered message can be more than a missed appointment.
How we build it at Catalizadora
At Catalizadora we build the agent with our MAGIA methodology: we map the real questions from your WhatsApp and your triage rules, design the conversation architecture, generate the agent in your practice's voice, implement it on your existing line, and leave it running autonomously.
Two things set us apart, and they matter especially in healthcare, where your patient data is sensitive:
- The code, the data, and the infrastructure are 100% yours. You don't rent software you depend on; the system is your asset.
- No retainers, no locked-in licenses. You pay for the build once. Operation is pass-through (hosting and tokens), on the order of 200 to 400 USD per month, with no margin from us.
The entry package, MAGIA Solo, costs 4,500 USD and we deliver it in 15 days. Compared to a year's salary for a second receptionist covering extended shifts, it pays for itself in a matter of months, and in the end the system is yours.
The next step
If your WhatsApp messages go cold every night while your receptionist sleeps, the first move isn't hiring another person: it's putting an agent on it that answers, qualifies, and books at any hour. Try it yourself: message our own AI agent on WhatsApp from the site and see how it responds. When you're ready to talk about your case, book a call with me at https://cal.com/pablo-estrada-hlqaql.