Skip to main content

AWS Lambda Function Integration

Fleak provides two methods for integrating AWS Lambda functions into your workflows: the AWS Lambda Function Node and LLM Function Call Integration. These options allow you to invoke Lambda functions directly from a workflow or have an LLM call a Lambda function to retrieve external data, giving you flexibility in leveraging AWS Lambda within Fleak.

AWS Lambda Function Node​

The AWS Lambda Function Node enables users to invoke an AWS Lambda function directly within their Fleak workflow without needing to set up an API Gateway. All that is required is an AWS connection with the necessary permissions.

For detailed info about the AWS Lambda Function Node, please refer here.

LLM Function Call with AWS Lambda​

In this integration, an AWS Lambda function can be attached to an LLM node, enabling the Large Language Model (LLM) to invoke the function when required. This setup is ideal for scenarios where the LLM needs to retrieve or process external data dynamically during interaction.

If you need to create a Lambda function in AWS before configuring it in Fleak, please visit the AWS Lambda Console.

Configuration Steps for LLM Function Call with AWS Lambda​

  • Define the Lambda Function in Fleak:

    • Go to the Functions page in Fleak and select New Function.
    • Provide the Lambda function's exact Name (case-sensitive), AWS Region, and AWS Connection details.

  • Function Call Definition: Describe the function’s purpose and define the required parameters using JSON Schema. This schema provides structure for the LLM, guiding it to format arguments correctly and return expected results.

Example JSON Schema for parameters:

{
  "name": "get_delivery_date",
  "description": "Returns a fixed delivery order date.",
  "parameters": {
    "type": "object",
    "properties": {},
    "required": []
  }
}

Visit here to learn more about LLM Function Calling.