What is the function of an "inference pipeline" in a data product context?

• A. It stores raw data.
• B. It executes data governance policies.
• C. It schedules batch jobs.
• D. It runs models to produce predictions or derived metrics for dashboards and decision-making assets.

Answer: (D)

An inference pipeline is the production path where a trained model is applied to new data to generate predictions or derived metrics that feed dashboards and decision-making assets. It takes in input data, runs the model to score or predict, and then post-processes the results into usable outputs that analytics apps, dashboards, or automated decisions can consume. This focuses on turning model outputs into actionable information, unlike simply storing raw data, enforcing governance policies, or scheduling batch jobs, which deal with data storage, policy enforcement, or task orchestration rather than generating model-based insights.