
Ultimate access to all questions.
You are working on a data pipeline for a customer churn prediction model in a telecommunications company. The pipeline processes data from customer interactions, billing information, and device usage. How would you design the pipeline to handle data from multiple sources and ensure accurate churn prediction?
A
Use a single processing engine for all data sources, without considering the specific characteristics of each source.
B
Design a modular processing architecture that can handle different data sources using specialized processing components and integrate the data for accurate churn prediction.
C
Process all data sources using a single data model, ignoring the unique characteristics of each source.
D
Use a batch processing approach to process the data in large batches, without considering the real-time requirements of the churn prediction model.