How generative models work, Microsoft Foundry, agents, retrieval and guardrails.
Sections from the books on generative ai and agents, free to read:
What happens between a prompt and a response, and how inference differs from training. AI-901
Which model setting controls randomness, which controls length and cost, and which ones are not set at deployment. AI-901
How to recognise each AI workload from a scenario. AI-901
Agents as model plus instructions, knowledge and tools, and the auto, required and none tool_choice values. AI-901
System messages, few-shot examples, chain of thought and grounding, and which fix suits which prompt problem. AI-103
File search, Azure AI Search, Bing grounding, function, OpenAPI, MCP and code interpreter tools compared. AI-103
How to keep humans in the loop and limit what an agent's tools can do. AI-103
Where guardrails check an agent run, what Prompt Shields catch, and when to block or annotate. AI-103
The indexer pipeline stages, built-in and custom skills, and knowledge store projections. AI-103
Vector fields and profiles, HNSW vs exhaustive KNN, hybrid queries with RRF, and the semantic ranker. AI-103
Prebuilt and custom analyzers, field extraction methods and Markdown output for RAG. AI-103
Speech to text, text to speech, translation, batch transcription and speaker recognition compared. AI-901
What Sora 2 can generate, its parameters and limits, and how the asynchronous jobs work. AI-103
Where prompts are processed for each deployment type, and which one meets data residency needs. AI-300
Search the series glossary for Foundry, or browse all 2,575 terms.
Know which AI fits the job, and why.
Know which service, tool and guardrail fits, and why.
Know how models reach production, and stay there.