You are building a machine learning model for your user behavior prediction problem using your company’s user interaction data stored in DynamoDB. You want to get your data into CSV format and load it into an S3 bucket so you can use it for your machine learning algorithm to give personalized recommendations to your users. Your data set needs to be updated automatically to produce real-time recommendations. Your business analysts also want to have the ability to run ad hoc queries on your data.
Which of the following architectures will be the most efficient way to achieve this?

Answer options:

A.Use AWS Data Pipeline to coordinate the following set of tasks: export DynamoDB data to S3 as JSON; Convert JSON to CSV; SageMaker model uses the data to produce real-time predictions; analysts use Amazon Athena to perform ad hoc queries against the CSV data in S3.
B.Create a custom classifier in an AWS Glue ETL job that extracts the DynamoDB data to CSV format on your S3 bucket; run your SageMaker model on the new data to produce real-time recommendations; analysts use Amazon Athena to perform ad hoc queries against the CSV data in S3.
C.Use AWS DMS to connect to your DynamoDB database and export the data to S3 in CSV format; run your SageMaker model on the new data to produce real-time recommendations; analysts use Amazon Athena to perform ad hoc queries against the CSV data in S3.
D.Use Kinesis Data Streams to receive the data from DynamoDB; use an ETL job running on an EC2 instance to consume the data and produce the CSV representation; run your SageMaker model on the new data to produce real-time recommendations; analysts use Amazon Athena to perform ad hoc queries against the CSV data in S3.

Answer: A
Option A is correct. AWS Data Pipeline is used here to schedule frequent runs of the described workflow: DynamoDB export, transformation, and running the model to give real-time predictions.
Option B is incorrect. This approach lacks the pipeline coordination described in
Option A.Option C is incorrect. AWS DMS does not support DynamoDB as a data source. Also, this approach lacks the pipeline coordination described in
Option A.Option D is incorrect. You would have to write more code to make this option work when compared to option A. You would need to write an extraction job to make the DynamoDB data into a Kinesis producer. You would also have to write the consumer ETL job. Also, this approach lacks the pipeline coordination described in
Option A.Reference:
Please see the AWS Data Pipeline developer guide titled What is AWS Data Pipeline, and AWS Database Migration Service user guide titled How AWS Data Migration Service Works specifically the section on sources, the Amazon Kinesis Data Streams developer guide titled Amazon Kinesis Data Streams Terminology and Concepts.
Here is a diagram of the best option: