Index

# 4. Data at Scale

Why This Matters¶

ML models are only as good as the data that feeds them. At Netflix scale, that means petabytes of viewing data, interaction logs, and catalog metadata processed through Spark and Hive pipelines. This section covers the data engineering skills that ML roles increasingly require.

Topics¶

• Apache Spark SQL
• Apache Hive
• Feature Stores & Pipelines
• Data for Post-Training