Microsoft AB-731: AI Transformation Leader Exam Prep

Explore the Microsoft 365 Copilot architecture from user prompts to Graph API context, plugins, and Azure OpenAI foundation models, ensuring privacy, compliance, and responsible AI.

Explore licensing options for Microsoft Copilot, including free Copilot web, pay as you go, and Microsoft 365 Copilot with graph grounding and apps integration. Understand admin setup and costs.

Explore in-app copilot settings for M365, including work and web modes and enterprise data protection. Configure memory and custom instructions to tailor responses with auto, quick, or think deeper modes.

Explore how the researcher agent in Microsoft Copilot for the 365 experience conducts in-depth research, analyzes user identity, policies, and email details to produce detailed summaries and policy mappings.

Analyze a sales dataset with the Copilot analyst agent to perform deep data analysis in a sandbox via Python, delivering revenue, average order value, regional and product insights, and charts.

Students learn to use Copilot in Word to visualize recipe text as tabular data, creating two-column tables for ingredients and amounts, and for directions, in a hands-on lab.

Learn to use Copilot in Excel online by formatting data as a table, querying insights, and generating pivot charts, outlier detection, and item and region counts in this hands-on lab.

Demonstrates creating a full PowerPoint deck from a Word document using Copilot for 365 with Graph grounding, including speaker notes, covering boiler versus furnace systems for commercial buildings.

Explore using Copilot in Outlook to draft emails, summarize longer messages, and generate professional replies. Learn coaching by Copilot to refine tone, add details, and tailor content for recipients.

In this hands-on lab, learn how Copilot in Teams summarizes chat threads, extracts key takeaways with citations, and surfaces follow-up tasks for efficiency.

Explore Microsoft Copilot Studio, a low-code Power Platform tool to build AI agents grounded in enterprise knowledge and publish them across teams and web.

Explore Microsoft Copilot Studio, sign in, and work in dev, staging, and production environments. Create embedded or standalone agents and flows via natural language or Power Automate.

Compare embedded and standalone Copilot Studio agents in Microsoft 365 Copilot. Embedded agents appear in same chat; standalone agents open a separate tab, leveraging enterprise knowledge beyond the 365 boundary.

Create a web grounded agent in Copilot Studio and configure knowledge sources, triggers, and prompts. Publish to Teams and Microsoft 365 Copilot, then test and iterate.

Explore how Microsoft Foundry solves agentic AI development challenges with a platform as a service, offering a 1000+ model catalog, agent service, and observability for scalable, responsible AI.

Deploy a Microsoft Foundry standalone project in the Azure Portal and explore Foundry Studio, the model catalog, the agent service, governance, and cost projection.

Deploy two models from the microsoft foundry model catalog—gpt 5.2 chat and claude opus 45—using serverless api inferencing, standard deployment, and content safety filters, then test in the chat playground.

Create your first agent in Microsoft Foundry with GPT 5.2, attach web search and code interpreter tools, set versions and guardrails, and review traces for debugging.

Explore Azure cognitive services and AI services, using rest APIs to add predictive capabilities to apps, and apply cognitive search, form recognizer, video indexer, and document intelligence for knowledge mining.

Experiment with Azure Document Intelligence by deploying a resource, using prebuilt analyzers for invoices and receipts, and extracting tables, key-value pairs, and entities via API calls and Document Intelligence Studio.

Explore the Azure AI language service and Azure Language Studio to deploy a language resource with default features like sentiment analysis and text summarization, and test PII extraction and redaction.

Explore the Azure computer vision resource in the portal by deploying a free tier instance, connecting it to Vision Studio, and testing features like face detection, liveliness, and portrait processing.

Azure ai content understanding uses cognitive services to turn unstructured data, text, images, video, and audio, into a structured, searchable dataset and supports building an analyzer and chatbots.

Explore Azure Content Understanding in the Microsoft Foundry hub, and analyze documents, audio, and video multimodal content to extract structured fields, OCR text, and transcripts with sentiment.

Explore retrieval augmented generation (rag) to ground chatbots in enterprise data using a retrieval pipeline, vector embeddings, and an augmented prompt, with Azure AI search support for multimodal data.

Fine tune pre-trained large language models with task specific prompt–response pairs to tailor behavior for business use cases, and compare this with retrieval augmented generation and practical costs.

Set up azure retrieval augmented generation infrastructure by provisioning blob storage, uploading pdf documents, enabling anonymous access, and deploying azure ai search with foundry models for multimodal vector retrieval.

Create an Azure AI search vector index from PDF documents in Azure Blob Storage for multimodal retrieval. Enable image verbalization with GPT 4.1 and Ada 002, and schedule incremental indexing.

Learn to use the Azure Machine Learning workspace to build, train, deploy, and manage your own AI and ML models with custom data, notebooks, AutoML, designer, pipelines, and endpoints.

Deploy and explore an Azure machine learning workspace, navigate to the studio, and use AutoML, notebooks, designer, and Prompt Flow to build, deploy, and monitor AI models at scale.

Learn how Microsoft Purview with Copilot enables unified data governance, applying sensitivity labels, data loss prevention, insider risk management, and eDiscovery across the M365 environment.

Explore how sensitivity labels in Microsoft Purview protect content across the M365 ecosystem and how Copilot respects and enforces these policies in SharePoint documents.