Choose Foundry models and services
AI-103 · Lesson 1 of 15 · 3:30
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AI-103 exam (Azure AI Apps and Agents Developer Associate). In this lesson: choosing models in the Foundry catalog (sold by Azure or by partners; large, small, reasoning, multimodal and embedding models; Foundry Local for offline use; leaderboards; deprecated GPT-4.1 and GPT-4o), the model router and Foundry Tools, RAG vs fine-tuning, grounding options (Azure AI Search tool, file search, web search, Foundry IQ), retrieval methods (keyword, vector, hybrid, semantic ranker, agentic retrieval), and memory, tools and knowledge for agents (memory store, toolbox, Agent Framework, routines). Includes three original practice questions.
Preview the topics
- Field technicians need an app that summarizes their notes on a laptop, often with no internet connection. What should you choose?
- Support tickets must have personal data, like phone numbers and addresses, detected and redacted, predictably and at low cost. What should you use?
- Several agents must share curated knowledge from SharePoint, Blob Storage and the web, and each user must only see what they’re allowed to. What should you use?