AI receptionist: how it works and how to set it up

An AI receptionist on the levelpartner.ai platform is the front desk for every inbound channel at once: phone calls, web chat, SMS and WhatsApp. It answers within seconds, recognises returning contacts, answers routine questions and books new enquiries into a calendar, at any hour. This guide treats the receptionist as one role spread across channels, because your prospects do not care which channel they used; they care how fast someone answered.

One front desk, several doors

Most firms staff their front door unevenly. The phone is answered during office hours. The website form waits until someone checks it. Social messages sit for days. Each door has a different response time, and the slowest one sets the impression for anyone who used it.

An AI receptionist removes that unevenness. A voice agent covers calls, a conversation agent covers chat and messaging, and both share the same knowledge base, the same calendar and the same contact records. A prospect who messages on Saturday and calls on Monday reaches a receptionist that already knows them.

What an AI receptionist does well

It greets quickly, gives accurate routine answers, collects details cleanly and books meetings. It never forgets to log a call. It works the same at 11pm as at 11am. What it does less well is handle ambiguity, emotion and anything that needs authority. Write your rules so those conversations reach a person fast.

Before you start

• Each inbound channel connected to your account: a phone number, the web chat widget, an SMS number and any social or WhatsApp accounts.
• A single knowledge base document covering who you serve, where you are, what you do and how a first meeting works.
• One booking calendar for new enquiries, with sensible availability and buffers.
• Clear hand-off rules and a named person for each type of hand-off.
• A short disclosure line saying the contact is speaking with an automated assistant.

How to set it up

1. Go to AI Agents > Conversation AI and click + Create Bot.
2. Select Guided Form Setup, then click Create New Bot.
3. Choose Appointment Booking as the bot type.
4. Select or create a Brand Voice.
5. Configure Bot Settings, then configure Bot Goals.
6. Save the bot and set its mode to Suggestive.
7. In AI Agents > Voice AI, open the Agent List tab.
8. Click + Create Agent, select Create Custom Agent, then click Continue.
9. Enter an Agent Name and select a Voice, Model and Language.
10. Write the prompt with the same qualifying questions as the bot.
11. Configure the Welcome Message and click Save.
12. Open Knowledge Base and select the same knowledge base.
13. Open the Deploy tab and select the phone number.
14. Under Call Routing, choose Use as Backup.
15. Save the deployment settings.

Why consistency matters more than speed alone

Speed gets the lead's attention. Consistency keeps it. If the chat receptionist says the first meeting is free and the phone receptionist says there is a fee, you have created doubt at the worst moment. A shared knowledge base and identical qualifying questions stop that. Review both agents together, not separately.

How much time it takes to run

A proper build across channels takes two to three days, mostly spent on the knowledge base and testing. After launch, plan an hour a week: read a sample of chats, listen to a few calls, update answers and check hand-offs were picked up. That hour is not optional; it is what keeps the receptionist accurate as your services and prices change.

When an AI receptionist is the wrong choice

If nearly all of your enquiries come through referral calls from people who expect to speak to a named partner, an automated front desk may feel wrong. Use it only for overflow. If your services need a regulated conversation before any booking, keep the receptionist to scheduling. And if routing calls to colleagues is your main need, the AI phone receptionist guide covers that job in detail. For booking-focused set-ups, see the AI booking assistant.

Troubleshooting

Different channels give different answers. The agents use different knowledge sources. Point both at the same knowledge base and remove channel-specific documents.

Bookings clash with existing meetings. The calendar is not connected to the staff member's external calendar. Connect it so busy times are blocked.

The same person appears as several contacts. Each channel created its own record. Ask for an email or phone number early and review your duplicate matching settings.

Hand-offs are missed at weekends. Nobody is assigned outside hours. Route weekend hand-offs to a callback booking rather than a person.

Frequently asked questions

Is an AI receptionist the same as a chatbot?

No. A chatbot usually covers one channel with fixed answers. An AI receptionist covers phone and messaging, answers from your knowledge base and books into your calendar. The chat bot with AI article explains the web chat side alone.

Can it take payments or deposits?

It can send a payment link from the platform if you set that up in a workflow. Keep pricing conversations simple and let a person handle anything negotiated.

What does it cost to run?

Both agents come with the licence. What you pay for is use: chat replies by AI token, voice time per minute plus the call itself, and each text or WhatsApp message sent. Volume drives the running cost, so start with the channels that matter most.

How do we keep its answers current?

Make one person responsible for the knowledge base and review it whenever services, fees or staff change. A monthly read-through catches anything that slipped.

The journey this guide describes