Skip to main content

Integrations

Not OnPrem

Integrations on the aiXplain platform allow you to securely connect external services - like Slack, Gmail, or Airtable, to your agents. These integrations power real-world task execution via natural language prompts.

List Available Integrations​

Use IntegrationFactory to explore all available integrations:

from aixplain.factories import IntegrationFactory

IntegrationFactory.list()
Show output

This helps you find available integrations like ZoomInfo, Slack, Notion, Airtable, and more.

Set Up an Integration​

  1. Go to Discover → Integrations tab.

  2. Search and select your desired integration.

Docusaurus themed imageDocusaurus themed image
  1. Click the integration card.
Docusaurus themed imageDocusaurus themed image
  1. Authenticate with your account.
Docusaurus themed imageDocusaurus themed image
  1. Once authenticated, the card updates to show your linked account.
Docusaurus themed imageDocusaurus themed image

Use the Integration Tool​

To use, call the tool with ToolFactory.

slack_tool = ToolFactory.get("67eff5c0e05614297caeef98")
slack_tool

Using Integrations in Agents​

Once your tool is ready, you can filter down to specific actions and pass the tool to an agent for execution.

Filter Actions​

slack_tool.action_scope = [action for action in slack_tool.actions if action.code == "SLACK_CHAT_POST_MESSAGE"]

Check Action Inputs​

Before using an action, you can check which fields are required:

action = slack_tool.action_scope[0]
slack_tool.get_action_inputs(action)

This will show you the required inputs like channel_id, text, etc., to properly execute the action.

Create and run an agent​

from aixplain.factories import AgentFactory

agent = AgentFactory.create(
    name="Test Agent",
    description="This agent sends messages to Slack.",
    instructions="You are a helpful assistant that can send messages to Slack.",
    llm_id="669a63646eb56306647e1091",
    tools=[
        slack_tool,
        AgentFactory.create_model_tool(model="6736411cf127849667606689")
    ],
)

response = agent.run("""Brazil, June 2nd, 2025
Tell Thiago on his Slack channel ('U022Q9LSLRM') the last AI news in the country""")
print(response.data.output)
Show output