As part of a pre-deployment readiness gate, an AI program undergoes a mandatory operational review. The review focuses on whether data entering the…

The scenario describes applying controls for data quality, formatting, standardization, cleansing, and removing restricted information before AI processing. This set of activities is the core function of the "Transform" stage in an Extract, Transform, Load (ETL) or Extract, Load, Transform (ELT) data pipeline. During transformation, raw data extracted from source systems is reshaped, cleaned, and enriched to meet the requirements of the downstream system—in this case, the AI environment. This stage ensures that data is compliant, consistent, and ready for consumption by AI models, directly addressing the requirements noted by leadership.

B. Load is the final stage where the transformed data is placed into the target system (e.g., a data warehouse or AI platform).

C. Extract is the initial stage responsible for retrieving raw data from various source systems; it does not involve cleansing or standardization.

D. Orchestrate refers to the management and scheduling of the entire pipeline workflow, not the specific data manipulation tasks themselves.

1. Han, J., Kamber, M., & Pei, J. (2012). Data Mining: Concepts and Techniques (3rd ed.). Morgan Kaufmann. (Chapter 3, "Data Preprocessing," details the steps of data cleaning, integration, and transformation, which are central to the Transform stage).

2. AWS. (2024). What is ETL?. AWS Documentation. Retrieved from https://aws.amazon.com/what-is-etl/. (Under the section "The ETL process," it explicitly states, "In the transformation step, the ETL tool applies a series of rules or functions to the extracted data to prepare it for loading into the end target").

3. Abadi, D. J. (2009). Lecture 1: Introduction to Data Warehousing. University of Maryland, Department of Computer Science. CMSC 724 Courseware. (Slide 21, "ETL," clearly defines the Transform step as involving data cleansing, reformatting, and combining).