Exam AWS Certified Machine Learning - Specialty topic 1 question 22 discussion

D is correct. Question has '"simple transformations, and some attributes will be combined" and Least development effort. Kinesis analytics can get data from Firehose, transform and write to S3 https://docs.aws.amazon.com/kinesisanalytics/latest/java/examples-s3.html

The best way to transform data is before it arrives to S3 so D should be best answer. But D is not completed. It should have another Firehose to deliver results to S3.

D. Insert an Amazon Kinesis Data Analytics stream downstream of the Kinesis Data Firehose stream that transforms raw record attributes into simple transformed values using SQL. Explanation: Since the data is already flowing through Amazon Kinesis Data Firehose, the least development effort solution is to use Amazon Kinesis Data Analytics, which supports SQL-based transformations on streaming data without requiring new infrastructure. Why is this the best choice? No major architectural changes – Data continues flowing from stores into Kinesis Data Firehose and then to Amazon S3. Simple SQL transformations – Since the changes are simple (e.g., attribute combinations), SQL is sufficient. Low operational overhead – No need to manage clusters or instances. Real-time processing – Transformed records immediately enter Amazon S3 for training.

Ans is D Amazon Kinesis Data Analytics provides a serverless option for real-time data processing using SQL queries. In this case, by inserting a Kinesis Data Analytics stream downstream of the Kinesis Data Firehose stream, the retail chain can easily perform the required simple transformations on the ingested purchasing records.

The best answer is to use a lambda, but the letter D can do it very good too in the absence of the lambda option.

I go with D. A tough question, though. And C are definitely out. They key to the question is that it does not say that the transformed data needs to be stored again in S3. It just needs to be sent to the model for training after being transformed. So a Kinesis Data Analytics stream is appropriate to do the transformation.

Legacy data -- Firehose -- Kinesis Analytics -- S3.This happens in near real time before the data ends up in S3. --Legacy data -- Firehose -- S3 is already happening (mentioned in first line in question), adding Kinesis Data Analytics to do simple transformation joins using SQL on the incoming data is the LEAST amount of work needed. Kinesis Data analytics can write o S3. here is the AWS link with working example.Even Though Udemy tutorial said it cannot write directly to S3 :) . https://docs.aws.amazon.com/kinesisanalytics/latest/java/examples-s3.html

It seems that LEAST developmnet effort: https://aws.amazon.com/fr/blogs/big-data/preprocessing-data-in-amazon-kinesis-analytics-with-aws-lambda/ and GRETAST development effort: https://aws.amazon.com/fr/blogs/big-data/optimizing-downstream-data-processing-with-amazon-kinesis-data-firehose-and-amazon-emr-running-apache-spark/

It's D https://aws.amazon.com/blogs/big-data/preprocessing-data-in-amazon-kinesis-analytics-with-aws-lambda/ "In some scenarios, you may need to enhance your streaming data with additional information, before you perform your SQL analysis. Kinesis Analytics gives you the ability to use data from Amazon S3 in your Kinesis Analytics application, using the Reference Data feature. However, you cannot use other data sources from within your SQL query."

It is D.

I think the answer is D, because require the LEAST amount of development effort.

it's D, kinesis analytic can easily connect with firehose

why not A. it seems good to me

D is correct...rest all need some kind of manual intervention as well as they are not simple..Firehose allows transformation as well as moving into S3

I think the answer is B. D would be correct if they didn't want to transform the legacy data from before the switch, but it seems like they do. Choosing D would mean that you'd have to use an EC2 instance or something else to transform the legacy data along with adding the Kinesis data analytics functionality. Also, there is no real-time requirement so daily transformation is fine.

I think the answer is B. D would be correct if they didn't want to transform the legacy data from before the switch, but it seems like they do. Choosing D would mean that you'd have to use an EC2 instance or something else to transform the legacy data along with adding the Kinesis data analytics functionality. Also, there is no real-time requirement so daily transformation is fine.

can be B