Ai fundamentals 0226

[perktime]

This pull request focuses on updating terminology and improving instructions across student and coach materials, as well as enhancing infrastructure provisioning for Azure Document Intelligence. The most important changes are the consistent replacement of "Azure Cognitive Search" with "Azure AI Search", improvements to local development environment setup instructions, and infrastructure enhancements for role assignments.

Terminology updates (Azure Cognitive Search → Azure AI Search):

• All references to "Azure Cognitive Search" in the coach guide (README.md) and solution notebook (CH-04-A-RAG_for_structured_data-Solution.ipynb) are updated to "Azure AI Search" for clarity and consistency. [1] [2] [3] [4] [5] [6]
• Related terminology in student challenge and learning resources is updated, e.g., "Azure Cognitive Services - What is Content Filtering" is now "Azure AI Services - What is Content Filtering".

Local development environment instructions:

• The student challenge instructions now recommend using Dev Containers for local development, provide step-by-step guidance, and clarify the risks of direct workstation setup. GitHub Copilot setup instructions are also added.
• Project verification steps are clarified, and redundant instructions about Microsoft Foundry resource creation are removed. [1] [2]

Infrastructure enhancements:

• The document.bicep module and main.bicep now accept and use a userObjectId parameter to assign the Cognitive Services Data Reader role to a specified user, improving resource access control. [1] [2] [3]

Student challenge improvements:

• Model router challenge instructions are expanded with categorized example prompts to help students observe model routing and cost optimization more clearly.

Minor content and formatting fixes:

• Several notebook cells are updated for clarity, including prompt engineering instructions and corrections to text encoding issues. [1] [2] [3] [4]This pull request updates the example code in the CH-06-AgenticAI.ipynb notebook to reflect recent changes in the agent API, ensuring compatibility and improving clarity in how multi-turn conversations are handled.

API usage updates:

• Replaced usage of agent.run_stream() with agent.run(..., stream=True) to align with the updated streaming interface. [1] [2]
• Updated multi-turn conversation logic to use agent.create_session() and pass the session parameter to agent.run, replacing the previous get_new_thread() and thread usage. This reflects the new session-based conversation management in the API.