You are a machine learning specialist working for a financial services firm. Your machine learning team is responsible for producing security master data for your quantitative analysis group. Your quant group uses your security master data to feed their quantitative analysis machine learning models to drive their stock selection for their active management portfolios. You need to stream security master data from various sources into your security master data store in near-real time. Which solutions meet your requirements in the most efficient manner? (Select TWO)

Answer options:

A.Stream your security master data using Kafka; extract the data into your online feature store using the PutRecord API call in small batch sets.
B.Stream your security master data using Spark Streaming; extract the data into your online feature store using the PutRecords API call in large batch sets.
C.Stream your security master data using Kinesis; extract the data into your online feature store using the PutRecord API call in small batch sets.
D.Stream your security master data using Apache Storm; extract the data into your online feature store using the PutRecords API call in small batch sets.
E.Stream your security master data using Kinesis; extract the data into your online feature store using the PutRecords API call in large batch sets.

Correct Answers: A and C
Option A is correct. You can use SageMaker Feature Store to house your security master data. You can ingest data into SageMaker Feature Store using the PutRecord API call using small batches of data. Kafka is a common service used to stream data into SageMaker Feature Store.
Option B is incorrect. The API call to put your data in small batches into SageMaker Feature Store is the PutRecord API call, not the PutRecords API call. Also, to meet your near-real time requirement, you should use small batches, not large batches of data.
Option C is correct. You can use SageMaker Feature Store to house your security master data. You can ingest data into SageMaker Feature Store using the PutRecord API call using small batches of data. Kinesis is a common service used to stream data into SageMaker Feature Store.
Option D is incorrect. The API call to put your data in small batches into SageMaker Feature Store is the PutRecord API call, not the PutRecords API call.
Option E is incorrect. The API call to put your data in small batches into SageMaker Feature Store is the PutRecord API call, not the PutRecords API call. Also, to meet your near-real time requirement, you should use small batches, not large batches of data.
Reference:
Please see the Amazon SageMaker developer guide titled Data Sources and Ingestion (https://docs.amazonaws.cn/en_us/sagemaker/latest/dg/feature-store-ingest-data.html), the Amazon SageMaker developer guide titled Get started with Amazon SageMaker Feature Store (https://docs.amazonaws.cn/en_us/sagemaker/latest/dg/feature-store-getting-started.html)