This week we covered the first stage of ML lifecycle: bringing raw data into AWS (collection) and storing it efficiently. This is the heart of MLA-C01 Domain 1 (Data Preparation, about 28%), and poor choices here topple downstream training costs and speed entirely.
Today we review by connecting S3, Kinesis, Glue, and Athena as one data pipeline. Rather than memorizing each service separately, understanding "how raw data becomes a trainable dataset" as a flow is more efficient for exam prep.
This week's services form a single flow.