Model Catalog and Strategic Selection
Exploration of Foundry’s versatility. Learning how to select, test, and deploy the right model for each business need, transitioning between elite models and efficient small-scale models.
Microsoft AI
Microsoft AI Foundry
4 - 6 hours 4 Modules
Overview
Main Theme: Microsoft AI / Copilot
Workload: 4 - 6 Hours
Level: Advanced
Target Audience: Solution Architects, IT Managers, and Engineering Leaders
ABOUT THE TRAINING
The Microsoft AI Foundry (formerly Azure AI Studio) is the ultimate command center for companies looking to develop robust AI applications. This training teaches how to orchestrate multiple large-scale language models, ensuring that innovation does not compromise the organization’s security or compliance. It is the ideal choice for teams operating in the Azure cloud and needing full sovereignty over their data.
Objective
Empower the technical leader to manage the complete lifecycle of an AI application, from choosing the ideal model to implementing rigorous security filters and integrating with secure corporate databases.
TRAINING MODULES
Content Safety and Compliance
Setting up layers of protection to prevent AI from generating toxic, biased responses or violating internal policies. The focus is on creating real-time security filters to protect the company’s reputation and ensure legal compliance.
Secure Integration with OneLake and Private Data
Techniques to connect AI models to corporate data stored in Azure in a fully private manner. The student will learn to use OneLake as the source of truth, allowing AI to respond based on real company information without ever exposing that data to the training of public models.
Governance, Costs, and Data Sovereignty
Operational management lab: how to control who utilizes resources, monitor costs, and ensure that data sovereignty is maintained. Discussion on the flexibility to switch models as the market evolves, without the need to rewrite the entire application infrastructure.