Booked stores calendars as booked_calendars, appointments as booked_appointments, and customers as booked_customers. Each appointment row links a customer to a calendar, a time slot, and a status, with custom field answers attached as appointment meta. SleekAI can be wired to all of that so the system message exposes the customer's upcoming appointments, the catalog of calendars and slot configurations, and the relevant custom field answers as named variables.

That changes what a returning customer can ask in chat. The bot can confirm the next appointment by calendar name, date, time, and status, answer slot-availability questions on a specific calendar, and reference the custom field answers the customer already filled in. None of those answers come from training data. They are direct reads on the Booked post types the calendar template already uses.

Display conditions scope the widget by page, role, or login state. Multibot keeps a public sales bot on service pages separate from a private support bot on the My Appointments page. Every conversation is logged inside WordPress so support leads can audit accuracy and refine the prompt over time.

Booked has powered small-team scheduling on WordPress for years, often for portfolio reviews, consultancies, and shared-resource calendars that mix several services on a single grid. The questions customers actually ask are precise. When is my appointment, did I tell you what to focus on, anything open next Friday morning, how do I cancel without losing the deposit.

A generic chatbot bolted onto a Booked site cannot see the calendars, has no idea what slot capacity remains, and will invent a focus area the customer never typed. That breaks trust faster than no chatbot at all. SleekAI flips the model.

The bot reads the same Booked post types and meta the calendar template reads, so its answer matches the page the customer is staring at. Per-page display conditions let a sales bot live on calendar pages with slot availability while a separate support bot lives on My Appointments with confirmed booking data, no leakage between them. The conversation log doubles as a quiet feedback loop on the calendar copy itself, surfacing the slot or service descriptions that customers misread on repeat.

Operators get back the time they used to spend confirming bookings by email, and customers get a scheduler that already knows who they are when they say hello.