Reference guides for the communication apps an AI agent can work with: every endpoint, the permission each one needs, and how to give agents safe, governed access.
The Aircall API is how an app or AI agent works with an Aircall phone system: listing and reading calls, tagging and commenting on them, managing the contacts shared across the workspace, and registering for live call events.
The Bird API is how an app or AI agent works with a Bird workspace: sending an SMS, WhatsApp, or email message over a channel, managing contacts and lists, running a conversation, or sending a one-time passcode to verify someone.
The Dialpad API is how an app or AI agent works with a Dialpad account: placing and transferring calls, sending an SMS, listing call history, and managing the users, contacts, and call centers behind a phone system.
The Discord API is how an app or AI agent works with a Discord server: posting messages to a channel, reading message history, managing channels and roles, and updating the members of a server.
The Front API is how an app or AI agent works with a shared inbox: reading and updating conversations, sending a reply, adding an internal comment for teammates, and managing contacts, tags, and channels.
The Gmail API is how an app or AI agent works with a Gmail or Google Workspace mailbox: reading and searching mail, sending and drafting messages, and organising what's there with labels.
The Google Chat API is how an app or AI agent works with a Google Chat workspace: listing and creating spaces, reading and posting messages, managing the people in a space, and adding reactions.
The Microsoft Outlook API is how an app or AI agent works with a user's Outlook mailbox: listing and reading messages, sending and replying to mail, creating and updating calendar events, and managing contacts.
The Microsoft Teams API is how an app or AI agent works with Microsoft Teams: listing teams and channels, reading and posting channel messages, reading chats and their messages, and managing team members.
The RingCentral API is how an app or AI agent works with a RingCentral account: sending text messages, listing call records, placing an outbound call, and reading a person's presence.
The Slack API is how an app or AI agent works with a Slack workspace: posting messages, reading channel history, creating channels, looking up people, and handling files.
The Telegram API is how an app or AI agent works through a Telegram bot: sending messages to a chat, receiving incoming updates, sending photos and documents, and managing members of a group.
The Twilio API is how an app or AI agent works with Twilio's communication channels: sending an SMS, MMS, or WhatsApp message, placing and controlling a phone call, buying or releasing a phone number, and managing the account and its subaccounts.
The Vonage API is how an app or AI agent reaches a phone network: sending an SMS or WhatsApp message, placing a voice call, verifying a user with a one-time code, or looking up a number.
The Webex API is how an app or AI agent works with a Webex workspace: posting messages to a space, creating spaces and adding people to them, looking up people, and scheduling meetings.
The WhatsApp Business API is how an app or AI agent works with a WhatsApp business account: sending text, template, and interactive messages to customers, marking incoming messages as read, uploading and fetching media, and managing message templates.
The Zoom API is how an app or AI agent works with a Zoom account: scheduling a meeting, listing and updating meetings, ending a running one, reading cloud recordings and transcripts, and pulling reports on who attended.
A communication API is how an app or AI agent sends and reads messages, calls, and meetings. The tools here, including Slack, Gmail, Twilio, and Zoom, carry email, chat, SMS, voice, and video, and their APIs let an agent post a message, send an email or text, read a conversation, or start a meeting.
Sending on a person’s or company’s behalf is a write action others can see, and reading a conversation can expose private messages, so the permission attached to an agent’s access is the main control. Access runs on an access token, usually scoped to particular actions like sending a message or reading a channel. A governed layer lets an agent that posts status updates write to one channel without being able to read every conversation in the workspace.
Bollard AI sits between a team's AI agents and the apps it runs on. Grant each agent exactly the access it needs, read or write, app by app, and every call is checked and logged.