5. Text, Speech, and Vision Solutions in Foundry

Phase 4 gave you the backbone: deploy a model, get a client, send input, read output. Phase 5 keeps that backbone and changes what flows through it. A text-analysis app sends text and reads structured results. A multimodal app sends an image alongside text. A speech app converts between audio and text. An image generator sends a prompt and gets back a picture. The exam tests whether you can match a modality to the right Foundry capability and recognize what the code is doing — so focus on what goes in, what comes out for each.