Booking a meeting can involve more than simply choosing a time. Teams may need to check availability, avoid double bookings, confirm time zones, create or update calendar events, notify the right people, send confirmation emails, and log the meeting for future reference.
This use case explores how a connected AI automation can turn a plain-language scheduling request into the right downstream action, whether that is checking availability, booking a meeting, rescheduling an event, sending a notification, or logging the details in a central system.
The Problem
Booking a meeting sounds simple until it is not. Checking real availability, avoiding double bookings, sending the right confirmation through the right channel, and remembering to log meeting details can involve several manual steps across chat, calendar, email, Slack, and spreadsheets.
When managed manually, these steps can create delays, inconsistent records, missed notifications, and confusion around availability or time zones. What should be a simple request can become a fragmented process across multiple tools.
The Solution
The AI Meeting Scheduling Assistant is a chat-based scheduling workflow built as a small team of specialised AI agents.
A user can make a request in plain language for availability, a booking, a reschedule, or a notification. An Orchestrator Agent then routes the request to the appropriate specialised agent, including a Calendar Agent, Notifier Agent, or CRM Logger Agent, each with the tools needed to complete the action.
How the Agent Works
- Receive a scheduling request. A user sends a message asking to check availability, book a meeting, reschedule an event, or notify someone about a meeting update.
- Identify the required action. The Orchestrator Agent interprets the request and routes it to the appropriate specialised agent rather than attempting to handle every task itself.
- Manage calendar activity. The Calendar Agent can check available time slots, create events, update existing bookings, or list meetings within a selected period.
- Send relevant notifications. The Notifier Agent can send Slack updates or email confirmations when a meeting is booked, changed, or cancelled.
- Log meeting information. The CRM Logger Agent records relevant meeting details in a central log, creating a lightweight meeting history for the team.
Technical Workflow
1. Chat entry point
A Chat Trigger node receives the user’s message and passes it straight to the Orchestrator Agent.
2. Strict single-tool routing
The Orchestrator’s system prompt is deliberately narrow: it must route every request to exactly one of three tool-agents, never explain its reasoning, and never answer the user directly.
For date-time requests, it is instructed to resolve relative terms (“today,” “tomorrow”) against a fixed reference date and pass everything to the Calendar Agent in Africa/Lagos ISO 8601 format.
3. Google Calendar Agent (availability & booking)
This sub-agent owns all calendar logic and is unusually detailed about time-zone handling. It must always operate in Africa/Lagos time, resolve “today” and “tomorrow” against $now, assume one-hour meetings unless told otherwise, and validate that start times are before end times and in the future.
It has four Google Calendar tools: check availability, create an event, update or reschedule an existing event, and list events in a date range. If a requested slot is busy, it should surface the actual conflicting meeting’s time back to the user.
4. Notifier Agent (Slack & Gmail)
A second sub-agent handles outbound notifications, telling a sales representative in Slack when a meeting is booked, rescheduled, or cancelled, and/or sending a confirmation email via Gmail.
Its prompt keeps it deliberately simple: deliver the message as given, without modifying it.
5. CRM Logger Agent (Google Sheets)
A third sub-agent appends meeting records to a “Meetings_log” Google Sheet, logging meeting ID, title, agenda, start/end time, attendee emails, notes, and whether it is linked to an associated contact.
This functions as a lightweight CRM and meeting history.
6. Shared reasoning and memory
A Think tool is available to the Orchestrator for working through more complex routing decisions before acting.
A single Simple Memory buffer, with a 30-message window, is shared across the Orchestrator and all three sub-agents, so context persists across a multi-step conversation even as control passes between agents.
Technology and Integrations
Built with: n8n, DeepSeek, Google Calendar, Slack, Gmail, Google Sheets, and a multi-agent architecture consisting of an Orchestrator Agent and three specialised sub-agents.
Outcome
The AI Meeting Scheduling Assistant creates a more structured way to manage calendar-related tasks.
A single chat message can trigger the appropriate action, whether that is checking availability, booking a meeting, rescheduling an event, sending a Slack notification, issuing an email confirmation, or logging the meeting in a central record.
The workflow reduces the need for users to move manually between calendar, messaging, email, and spreadsheet tools while keeping meeting information and communication more organised.
Explore Custom AI Automation for Your Business
The AI Meeting Scheduling Assistant is one example of how AI automation can support more efficient scheduling and operational workflows.
Divverse Labs designs and builds custom AI agents, automation workflows, internal tools, and connected systems for a wide range of business processes.
From sales, marketing, recruitment, customer support, and reporting to finance, operations, and internal team workflows, each solution is designed around the way your business works.
Request a Quote for a Custom AI Automation Solution to get started.