Develop custom copilots with Azure AI Foundry

This training is intended for: 

• Data scientists who have knowledge of generative AI models and Python, and want to create, customize, and deploy their own Copilots. 

• AI engineers who have experience working with generative AI models and Python, and want to enhance their skills in developing custom Copilots. 

1.Plan and prepare to develop AI solutions on Azure 

• Introduction 

• What is AI? 

• Azure AI services 

• Azure AI Foundry 

• Developer tools and SDKs 

• Responsible AI 

• Exercise – Explore the Azure AI Foundry portal 

2.Explore and deploy models from the model catalog in Azure AI Foundry portal 

• Introduction 

• Explore the language models in the model catalog 

• Deploy a model to an endpoint 

• Improve the performance of a language model 

• Exercise – Explore, deploy, and chat with language models 

3.Get started with prompt flow to develop language model apps in the Azure AI Foundry 

• Introduction 

• Understand the development lifecycle of a large language model (LLM) app 

• Understand core components and explore flow types 

• Explore connections and runtimes 

• Explore variants and monitoring options 

• Exercise – Get started with prompt flow 

4.Build a RAG-based agent with your own data using Azure AI Foundry 

• Introduction 

• Understand how to ground your language model 

• Make your data searchable 

• Build an agent with prompt flow 

• Exercise – Create a custom agent that uses your own data 

5.Fine-tune a language model with Azure AI Foundry 

• Introduction 

• Understand when to fine-tune a language model 

• Prepare your data to fine-tune a chat completion model 

• Explore fine-tuning language models in Azure AI Studio 

• Exercise – Fine-tune a foundation model 

6.Evaluate the performance of generative AI apps with Azure AI Foundry 

• Introduction 

• Assess the model performance 

• Manually evaluate the performance of a model 

• Assess the performance of your generative AI apps 

• Exercise – Evaluate the performance of your generative AI app 

7.Responsible generative AI 

• Introduction 

• Plan a responsible generative AI solution 

• Identify potential harms 

• Measure potential harms 

• Mitigate potential harms 

• Operate a responsible generative AI solution 

• Exercise – Explore content filters in Azure AI Studio 

• Understanding of Azure AI services (AI-900) 

• Basic Python programming skills (PYTH01) 

• Copilot Foundations (AI-3018)