How to Build an AI Receptionist with Retell AI (Beginner Tutorial)

Why This Tutorial Exists

Most "Retell AI tutorial" content stops at "sign up and pick a voice." That is not enough to charge a local business 297 dollars a month. This walkthrough takes you from an empty Retell account to a working AI phone receptionist that answers, qualifies, and books appointments — in under an hour.

If you run a marketing or automation agency (or want to), this is the exact build you resell. The end client is a dentist, an HVAC company, or a law firm that misses 30 to 40 percent of inbound calls. Your job is to catch those calls with an agent that sounds human and books work. Retell AI is the platform most agencies land on because it has the best latency, a clean agency sub-account model, and native calendar booking.

We will use a dental office as the running example because the call patterns are predictable and easy to template across clients.

What You Need Before You Start

You can have all of this in 15 minutes:

• A Retell AI account (free to start, pay per minute after)
• A Twilio account for the phone number (about 1 dollar a month per local number)
• A Cal.com or Google Calendar for booking (Cal.com is free and integrates cleanly)
• An LLM key is not required — Retell bundles model access into the per-minute price

That is the entire stack. No code, no servers, no Make.com middleware. Everything below happens inside the Retell dashboard.

Step 1: Create the Agent

Inside Retell, go to Agents and create a new conversational agent. You will pick three things up front:

Setting

Recommended choice

Why

|---------|-------------------|-----|

LLM

GPT-4o mini or Claude Haiku

Fast and cheap enough for high call volume; upgrade to Sonnet only if reasoning is weak

Voice

Cartesia Sonic (bundled)

Sub-second latency, natural tone, no extra ElevenLabs bill

Language

English (US)

Match your client's market; Retell supports multilingual if needed

Name the agent something you can find later across many clients, like "Bright Smiles Dental — Reception." When you run 10 clients, naming discipline saves you.

Step 2: Write the System Prompt

This is where 90 percent of the quality lives, and where most beginners get it wrong. A good receptionist prompt has five parts: identity, scope, knowledge, behavior rules, and a booking instruction.

Here is the structure to fill in:

Identity. You are the front desk assistant for Bright Smiles Dental in Austin, Texas. You are warm, brief, and efficient. If asked, you confirm you are an AI assistant for the practice.

Scope. You answer questions about hours, location, services, insurance, and new-patient onboarding. You book, reschedule, and cancel appointments. You do not give medical or clinical advice.

Knowledge. Hours: Monday to Friday, 8am to 5pm. New patient exams are 297 dollars or covered by most PPO plans. We do not currently accept Medicaid. Parking is free in the rear lot.

Behavior rules. Keep replies under two sentences. Ask one question at a time. If the caller is angry or describes a dental emergency with severe pain or bleeding, say you will connect them to staff and trigger a transfer. Never invent prices or insurance details — if you do not know, offer to take a message.

Booking instruction. When the caller wants an appointment, collect their full name, phone number, and preferred day, then check availability and confirm.

Keep the whole prompt under 500 words. Long prompts slow the agent down and make it hallucinate. Tight scope plus real knowledge beats a wall of text every time.

This prompt scaffold is the single biggest time sink when you onboard clients, which is exactly why a ready-made [prompt library](/shop) and [setup blueprint](/shop) pay for themselves on your first build.

Step 3: Set the Greeting and Conversation Flow

Retell lets you set the opening line the agent speaks. Make it disclose the business and invite a request in one breath:

Thank you for calling Bright Smiles Dental, this is the front desk assistant — how can I help you today?

That opener does double duty. It sounds professional, and it satisfies the AI-disclosure expectation in states like California where identifying as an automated assistant matters. For inbound calls this is a soft, friendly disclosure rather than a legal warning, and it keeps the conversation natural.

Set the interruption sensitivity to medium so the agent stops talking when the caller jumps in. A receptionist that talks over people sounds robotic instantly.

Step 4: Connect Calendar Booking

This is the feature that turns a toy into a paid service. A receptionist that just answers questions is worth maybe 99 dollars a month. One that books appointments straight into the client's calendar is worth 297 to 497.

In Retell, add a function (sometimes labeled a custom tool or action) and connect it to Cal.com or Google Calendar:

In Cal.com, create an API key and grab the event type ID for "New Patient Exam"

In Retell, add the calendar booking function and paste the key

Tell the agent in the prompt when to call it — for example, after collecting name, phone, and preferred time

Add a second function for checking availability so the agent does not offer slots that are already full

Now when a caller says "can I come in Thursday morning," the agent checks real availability, offers two open slots, and writes the booking. That is the demo that closes clients.

Step 5: Attach a Phone Number and Test

Buy a local Twilio number that matches the client's area code (callers trust local numbers) and import it into Retell, or buy a number directly inside Retell. Assign it to your agent.

Before you ever go live, call the number yourself and run these scenarios:

• A normal booking ("I want a cleaning next week")
• A pricing question ("how much is a new patient exam")
• An out-of-scope question ("can you tell me if my tooth needs a root canal")
• An interruption (start talking over the agent mid-sentence)
• An emergency phrase ("my mouth is bleeding and I'm in a lot of pain")

You are listening for three things: latency under one second, no invented facts, and a clean handoff or message when the agent hits its limits. Expect to revise the prompt two or three times. That is normal — the first version is never the deployed version.

Step 6: Go Live and Monitor

Point the client's existing business line to the Twilio number using call forwarding, or set the agent to handle overflow and after-hours only if the client is nervous. After-hours-only is a great low-risk way to land a skeptical first client.

Watch the first 20 to 30 calls in the Retell call log. Every transcript shows you where the prompt is weak. Within the first week you will have a tight, accurate agent.

What This Actually Costs You to Run

Per client, monthly, at roughly 500 minutes of calls:

Item

Cost

|------|------|

Retell AI (about 0.07 per minute)

35 dollars

Twilio local number

1 dollar

Cal.com

Free tier

Your cost

about 36 dollars

Charge the client 297 dollars a month and your gross margin is north of 85 percent. Ten clients is roughly 2,610 dollars a month in margin on about an hour of monthly maintenance each. That is the math that makes AI receptionists the cleanest service an agency can resell in 2026.

Common Mistakes That Sink Beginner Builds

• Bloated prompts. Over 500 words and the agent slows down and starts inventing answers. Cut ruthlessly.
• No "I don't know" path. Always give the agent a fallback (take a message, offer a callback) so it never guesses on prices or insurance.
• Skipping the test calls. Going live without running edge cases is how you lose a client in week one.
• Premium voice you don't need. Cartesia is bundled and excellent. Do not add an ElevenLabs bill until a client specifically asks for voice cloning.
• Outbound calling without consent. Inbound is the safe zone. Outbound triggers TCPA rules — do not freelance there.

Where to Go From Here

You now have a working AI receptionist and the cost math to price it. The slow part is doing it again for the next client — rewriting prompts, rebuilding booking flows, and packaging the offer.

If you want the shortcut, the [AI Receptionist Agency Launch System](/start) gives you the done-for-you version of this build: the four guides that cover client acquisition and pricing, plus the templates you reuse on every deployment. The standalone [prompt library and setup blueprint](/shop) drop straight into the steps above so your second build takes 20 minutes instead of an hour. And if you want a deeper feature comparison before you commit to the platform, our full [Retell AI review](/tools/retell-ai) breaks down latency, the agency sub-account model, and pricing tiers.

Build one, demo it, charge for it. The platform is ready — the only thing between you and your first paying client is the offer.