Get started with Generative AI & Agents in Azure - Foundry Models, Tools & Knowledge Sources #ai901

17 Просмотры

Спасибо! Поделитесь с друзьями!

Вам не понравилось видео. Спасибо за то что поделились своим мнением!

Издатель Sep 1, 2026

Generative AI can write, code, and create — but on its own, it can't take action or remember anything between calls. That's where agents come in, and in this video we break down exactly how Microsoft Foundry turns a plain generative AI model into a reusable agent with tools and knowledge of its own. This is a core part of the AI-901 exam objectives, so we cover it in full, from the model catalog all the way to calling a live agent from Python.

In this video we cover:
✅ What generative AI is, and five real-world use cases you'll actually recognize
✅ How the Microsoft Foundry model catalog works, and what each model listing tells you
✅ The most common model families — GPT-5.x, Claude Opus 4.5, Mistral Large 3, and GPT-4.1
✅ How to evaluate models using benchmarks, leaderboards, and quality metrics
✅ What actually happens when you deploy a model, including tokens, TPM, and throttling
✅ Using the Foundry Playground to test prompts before writing any code
✅ Building a lightweight Python chat client with the Responses API
✅ The real difference between a model and an agent
✅ Creating an agent in Foundry with a model, instructions, and tools
✅ Grounding an agent with knowledge using retrieval-augmented generation
✅ Calling a saved agent from your own client application

Join this channel to get access to perks:
https://www.youtube.com/channel/UCtmTEekMEUwPIcgOFaRLpqw/join