Which component is responsible for applying a model to live production data and returning inference results?

• A. Data Lake
• B. Data Exploration and Training Platform
• C. Vector Database
• D. AI System Production

Answer: (D)

The question is about where the model actually runs on new, real-time data to produce predictions. That happens in the production-serving component of an AI system—the AI System Production. This part is designed to handle live data streams or batched requests, perform any necessary preprocessing, run the trained model to generate inference results, and deliver those results to downstream systems or users. It also manages concerns like latency, scaling, model versioning, and monitoring.

The other components play different roles in the lifecycle: the data lake stores raw and processed data for analysis; the data exploration and training platform is where models are built, experimented with, and trained; the vector database stores embeddings to support fast similarity searches and retrieval during some inference workflows, but it does not by itself apply the model to live data and return inference results.