AWS Rekognition for 'rekognizing' images.
'Hello World' was the gentle dipping of the toe for learning programming, a baton picked up by image recognition for learning (or starting to learn) AI,
the AWS way.
First: The AWS Architecture.

Second: The Steps.
- Create a bucket

- Let ACLs be disabled because we will be using a policy to control access

- Leave all other settings default.
- Voila... you have an S3 bucket

- Put a photo of your choice in the S3 bucket.

- Create a lambda function using Python BUT before that, create the policy below and map it to a role. The function should be allowed to run under either an existing role (or you can create a new if you want).
- The role you make (or edit) should have the following policies:
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Sid": "DynamodbLimitedAccess",
"Action": [
"rekognition:Get*",
"rekognition:List*",
"rekognition:DetectLabels"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Sid": "SecretsManagerLimitedAccess",
"Action": [
"s3:Describe*",
"s3:Get*",
"s3:List*"
],
"Resource": "*"
}
]
}
- Add the Python code below to the Lambda
import json
import boto3
def lambda_handler(event, context):
bucket_name = "rekognition2026"
image_obj_name = "Sample-png-image-1mb.png"
try:
rkClient = boto3.client("rekognition", region_name="us-east-1")
try:
rkResponse = rkClient.detect_labels(
Image={
'S3Object': {
'Bucket': bucket_name,
'Name': image_obj_name
}
},
)
print(rkResponse['Labels'])
return rkResponse['Labels']
except Exception as e:
print("Get labels failed because ", e)
except Exception as e:
print("Client connection to Rekognition failed because ", e)
- Deploy code
- Test it

I write to remember, and if, in the process, I can help someone learn about Containers, Orchestration (Docker Compose, Kubernetes), GitOps, DevSecOps, VR/AR, Architecture, and Data Management, that is just icing on the cake.