How to write a lambda function for the Map task in the sagemaker?

Question

I have a machine learning model deployed in SageMaker, and I'm currently working on a process to handle multiple inputs for predictions. I understand that AWS Lambda can be used to implement a Map task to process these inputs in parallel and continuously (in csv format), but I'm not sure how to write the Lambda function for this specific purpose. My Question: Could someone guide me on how to write a Lambda function that can take multiple inputs in csv format, to interact with a SageMaker endpoint? I have chosen my csv file in s3 bucket(as a part of map task to be used in the lambda function) and I have left the test event space empty.

Below is the syntax I've written so far but I'm not sure about the section where we send the request to the sagemaker endpoint. Thank you in advance.

import json import boto3

My deployed model's endpoint name

ENDPOINT = 'my-endpoint-name'

def lambda_handler(event, context): # Assuming event contains CSV data as a string csv_data = event['csv']

# Parsing the CSV data to extract the image name
# This depends on your CSV format. Here's a simple example:
lines = csv_data.split('\n')
first_line = lines[0]
image_name = first_line.split(',')[0]  # Assuming image name is in the first column

# Now image_name contains the name of the image to be processed
# Add your logic here to use image_name as needed, for example, sending it to the SageMaker endpoint

***# Example of sending a request to the SageMaker endpoint (Is this correct?)
response = runtime.invoke_endpoint(EndpointName=ENDPOINT,
                                   ContentType='text/plain',  # Change if needed
                                   Body=image_name)***

# Process the response from the endpoint
predictions = json.loads(response['Body'].read().decode())

return {
    'statusCode': 200,
    'Predictions': predictions
}