2. Develop AI Solutions by Using Azure Data Management Services (25–30%)

This is the exam's heaviest domain, and it implements the vector-search principle from Section 1.1.2 three different ways. The data layer of the anatomy from 1.1.1 splits into three engines with three distinct jobs: Cosmos DB for globally scaled operational data, PostgreSQL for relational data with vectors, and Redis for the hot path. Expect scenario questions that test whether you know each engine's cost model, its vector capabilities, and its operational sharp edges.