Every Microsoft exam we cover

Spotting, investigating and shutting down attacks using Microsoft Defender and Sentinel.

Controlling who can sign in, what they can reach, and for how long.

Protecting data with Microsoft Purview: labelling, DLP, retention and insider risk.

Securing Azure identity, storage, networking, compute and AI workloads, and watching posture.

Designing a security strategy an organisation can actually run.

Securing code, secrets and dependencies on GitHub: what the security suites cover and where they overlap, secret protection and push protection, the dependency graph and Dependabot, CodeQL scanning and triage, prioritising and remediating at scale, and rolling the whole thing out across an enterprise.

The day to day of running Azure: identity, storage, networking, compute.

Choosing between Azure services when the requirements conflict: identity, data, continuity and infrastructure.

Azure networking end to end: addressing and routing, hybrid connectivity, delivery, private access and network security.

Azure DevOps and GitHub together: branching and traceability, YAML pipelines and deployment strategies, secrets and scanning, and telemetry.

Azure Virtual Desktop end to end: host pools, session hosts and images, the network and storage underneath them, identity and session host security, FSLogix profiles and app delivery, and the monitoring, scaling and recovery that keep a deployment running.

Automating with GitHub Actions: triggers, jobs and matrices, passing data between steps, reading a failed run, reusable workflows against starter workflows against composite actions, writing and publishing your own actions, runners and secrets at enterprise scale, and the security and cost decisions underneath all of it.

Building and running agents that do real development work: separating planning from execution, tools and MCP servers with scoped permissions, agents in repositories and CI, memory and state across long tasks, evaluating and tuning agent behaviour from failures, coordinating several agents without them colliding, and the autonomy levels and guardrails that decide what an agent may do unsupervised.

Running GitHub at enterprise scale: managed users against personal accounts, SAML, SCIM and team synchronisation, the deployment and licensing models, enterprise policies and rulesets, audit logging, tokens and app approval, Actions administration including runner groups and secret scope, and reading the usage that all of it produces.

Running Windows Server in a hybrid estate: domain controllers, trusts, sites and replication, Group Policy, remote management and Azure Arc, Hyper-V and virtual machines in Azure, name resolution and addressing, file services from Storage Spaces to Azure File Sync, hardening the operating system and the directory, and the monitoring and troubleshooting that keeps it all diagnosable.

Running SAP on Azure: sizing and migration strategy for NetWeaver, Business Suite and HANA, landing zones and integration with RISE with SAP, certified virtual machine types, proximity placement groups and accelerated networking for latency, NetApp Files and Write Accelerator for storage, Pacemaker clustering with fencing for Central Services and HANA, site recovery and backup that meet an SAP service level, and the monitoring, automation and cost work that keeps a landscape running afterwards.

Running Teams properly: governance, external access, meetings, calling, and the reasons things break.

Managing and securing endpoints with Intune: enrolment, Autopilot, policies, updates and apps.

Teams as a telephone system and a room: PSTN connectivity from Calling Plans to Operator Connect to Direct Routing, auto attendants and call queues, emergency calling and location services, meeting and webinar policies, audio conferencing, and the rooms, devices and network quality that decide whether any of it works.

Administering Microsoft 365 alongside the AI services now built into it: tenant configuration with branding, domains, organisational settings, licensing for Copilot, Agent 365 and Copilot Studio, Microsoft 365 Backup, network connectivity insights and service health, then the workloads themselves through mailboxes, teams, SharePoint sites and the settings that decide what Copilot may reach. Governance and security across identity in Entra with users, guests, groups, roles, privileged identity management, administrative units and bulk work through Graph PowerShell, authentication methods, password protection, self-service reset and Conditional Access with identity protection, Defender for Office threat policies, investigation and attack simulation, and Purview for data loss prevention, information protection, lifecycle management and data security posture across AI activity. Then the AI services proper: Copilot readiness and the oversharing that blocks it, search and web settings, tenant Copilot configuration, Cowork, third-party providers and connectors, agents through Entra Agent ID, access packages, the agent registry and the tools in Agent 365, and the monitoring of cost, usage, adoption and health through the Copilot Control System.

Building analytics in Fabric: lakehouses and warehouses, preparing data, and semantic models that perform.

Power BI end to end: shaping data in Power Query, modelling and DAX, building reports that read well, and governing what you publish.

Building on Azure AI Foundry: choosing models, retrieval and grounding, agents and their tools, safety and evaluation, and the vision, language and extraction services around them.

Data engineering in Fabric: workspaces and access control, ingestion and transformation across pipelines, dataflows, notebooks and KQL, streaming, and the monitoring and tuning that keeps it running.

Working with GitHub Copilot properly: responsible use and its limits, the features across the IDE, CLI and github.com, how data flows through it, prompt crafting, everyday productivity, and the privacy settings and content exclusions an organisation has to get right.

Running SQL on Azure and on-premises: choosing between Database, Managed Instance and SQL on virtual machines, migrating onto them, securing them, monitoring and tuning what they do, automating the routine work, and the high availability and recovery design underneath all of it.

Developing in SQL rather than administering it: table and index design including specialised and JSON columns, advanced T-SQL from window functions to regular expressions and graph queries, security and query performance from the developer's seat, database projects and deployment pipelines, and the AI features now inside the engine, vectors, embeddings, hybrid search and retrieval augmented generation.

Engineering data on Azure Databricks: choosing and configuring compute, organising and securing Unity Catalog with privileges, masks and lineage, ingestion through Auto Loader, declarative pipelines and streaming, table design from clustering to slowly changing dimensions, and the jobs, bundles, monitoring and Spark tuning that keep it running.

Operating machine learning and generative AI in production: workspaces, datastores, compute and assets, experiment tracking and training pipelines, model registration and endpoints with safe rollout, drift and performance monitoring, Foundry environments and model deployment, evaluation and observability for generative applications, and the retrieval and fine-tuning work that keeps them accurate.

Building AI applications on Azure: containers in Container Registry, App Service, Container Apps and Kubernetes, the data services behind retrieval including Cosmos DB, PostgreSQL and Redis with vectors, messaging and Functions, and the secrets, configuration and tracing that make it operable.

Designing and running systems where several agents work together: decomposing a goal into agents and tools, orchestration patterns and agent-to-agent protocols, memory and context shared across a team of agents, retrieval and tool ecosystems including MCP servers, evaluation and tracing of agent behaviour, guardrails at every interception point, per-agent identity and authentication flows, and the release strategies that put a fleet of agents into production.

Building agents in Copilot Studio and connecting them to everything else: planning identity, channels and reusable components, agent flows with human approval and error handling, topics with custom prompts, adaptive cards and variables, enterprise knowledge through connectors and Azure AI Search, tools from custom connectors, REST APIs, protocol servers and computer use, multi-agent collaboration with Foundry agents, Fabric data agents and the agent-to-agent protocol, and the testing, evaluation and solution lifecycle work around it.

Architecting agentic solutions across Microsoft business applications: assessing where agents belong and whether the grounding data is fit for them, the AI adoption path from the Cloud Adoption Framework, multi-agent designs across Microsoft 365 Copilot, Copilot Studio and Foundry, prebuilt against custom agents and when a custom model is warranted, prompt libraries and model routing, return on investment and build against buy, designing topics, knowledge, triggers, agent flows and prompt actions in Copilot Studio, extensibility through connectors, Model Context Protocol and Computer Use, orchestrating the AI already in Dynamics 365, then monitoring, telemetry, testing, lifecycle management for agents, data and models, and the security, governance, responsible AI, residency and prompt-manipulation work that has to hold all of it together.

Configuring AI-enhanced selling in Dynamics 365 Sales: the core setup of mailboxes, business process flows, the timeline, data import, security and the Microsoft 365 integrations, product families, price lists and currencies, then the intelligence layer of the Sales accelerator, sequences and work assignment, conversation intelligence, predictive scoring, relationship intelligence, Copilot summarisation, forecasts and goals, lead qualification with the Sales Qualification Agent in research-only and research-and-engage modes, opportunity work with the Sales Opportunity, Close and Research agents and the research canvas, opportunity products, pricing and the pipeline view, and extending it all through the mobile app, Teams calling, an SMS channel, Power Automate flows and embedded Power Apps and Power BI content.

Engineering a contact centre on Dynamics 365: deployment through the Copilot Service workspace, embedded mode, connectors, the Agent hub, solutions and lifecycle management, users, roles and capacity, then the channels themselves, which are a third of the exam: chat widgets, proactive chat, pre-chat surveys, record-based and custom messaging channels, the live chat and messaging SDKs, and the voice channel with phone numbers, inbound and outbound calling profiles, third-party interactive voice response, call recording, transcription and translation, plus context variables, masking, automatic identification, consult and transfer, proactive outbound campaigns and their dialling methods, and workforce management with forecasting and shifts. Then voice agents in Copilot Studio with dual-tone, speech and generative orchestration, session initiation protocol header transfers and multilingual behaviour, work distribution through queues, intent-based, preferred-representative and AI-enabled skills routing with conversation diagnostics, the representative experience with custom productivity panels and the Omnichannel and app profile manager interfaces, and analytics from the supervisor app through the embedded Power BI editor to Application Insights telemetry.

Building AI-enabled solutions on Power Platform with low code: analysing a requirement into apps, flows and agents, evaluating built-in agents and extensibility, choosing environment types and a lifecycle strategy with solutions and pipelines, modelling data in the data workspace with tables, columns, relationships, prompt columns, row summaries, views, forms and security, then building model-driven apps with generative pages, charts and dashboards and canvas apps with delegation-aware data access, accessibility, reusable components, named formulas, user-defined functions, variables and collections, error handling and Monitor, and finally the automation and intelligence: cloud flow triggers, connectors, approvals, conditions and loops, testing and troubleshooting, prompts built in the AI hub from templates or blank with knowledge, inputs and model settings and consumed from both apps and flows, prebuilt models, and the business logic layer of rules, process flows and calculated, rollup and formula columns.

Developing for Dynamics 365 finance and operations apps: the cloud and on-premises components and lifecycle management through Lifecycle Services and the Implementation portal, Visual Studio with extension models, Application Explorer and the element designers, Azure DevOps version control and continuous delivery, extending forms, menus, tables, views, queries, maps, data entities and classes, X++ from its structured constructs and data access through to inheritance, attributes, query objects, Chain of Command, delegates and event handlers, the SysOperation, workflow and SysExtensionSerializer frameworks, unit testing, reporting with SSRS, Power BI, Excel and Electronic reporting, workspaces and key performance indicators, integration through OData, custom services, business events, dual-write and virtual entities, data management projects and change tracking, and security with duties, privileges and Extensible Data Security alongside caching, temporary tables, set-based queries, indexes and trace analysis.

Implementing Business Central for a company: assisted setup, configuration packages and data migration, users, permission sets and security filters, dimensions and approval workflows, the whole financial configuration from general ledger and posting groups to payables, receivables and fixed assets, inventory, costing and pricing, and then the daily operations, quotes to orders to invoices, journals, payments, reconciliation and inventory transactions.

Extending Business Central in AL: the extension model and what the base and system apps are for, setting up the development environment and managing configuration and workspaces, building tables, pages, reports, XMLports, codeunits, permission sets and queries, the language itself with its data types, procedures, error handling and access modifiers, the user assistance and onboarding model, testing with the test toolkit, telemetry, and integration through HTTP, JSON, APIs and bound actions.

Implementing Dynamics 365 Finance: the chart of accounts, financial dimensions, account structures and advanced rules, ledgers, currencies and posting layers, journals with their controls and intercompany posting, cash and bank including advanced reconciliation, period close, consolidation and ledger settlement, tax, then receivables with credit, collections and subscription billing, payables and expense management, budgeting with control and planning, and fixed assets from books and depreciation profiles through to disposal.

Implementing Dynamics 365 Customer Service: cases and the rules that create and update them, the resolution experience, security roles and the timeline control, knowledge management with internal and external sources, translation and Copilot-led authoring, Teams collaboration and suggested contacts, the Copilot features from Ask a Question to Draft a Response and the Case Management Agent, service level agreements with their items, KPIs and timers, unified routing with workstreams, queues, capacity profiles, work classification and skills-based routing, productivity tools including scripts, slugs and macros, the Copilot Service workspace with session templates, experience profiles, the Inbox and tab templates, and the Dataverse work underneath: forms, views, columns, tables, relationships, search, notifications, email templates and Customer Voice surveys.

Implementing Dynamics 365 Supply Chain Management: products, variants, bills of materials and the dimension, tracking and reservation groups underneath them, costing versions and trade agreements, inventory journals, transfer orders, closing, quality orders and asset maintenance, procurement from requisition through to vendor collaboration and landed cost, sales agreements, rebates and availability promises, warehouse configuration with location directives, work and wave templates and the mobile app, transportation with carriers, routes and freight reconciliation, and master planning with coverage groups, fences and safety stock.

The vocabulary of Microsoft security, for people starting out.

What Azure is, how it is organised, and what the main services are for.

What AI can do on Azure, and building simple solutions with Microsoft Foundry.

The vocabulary of data on Azure: structured against unstructured, transactional against analytical, the SQL family, Cosmos DB, and what an analytics pipeline is made of.

What Power Platform is for: Dataverse and connectors, canvas and model-driven apps, cloud and desktop flows, and building agents in Copilot Studio.

How GitHub actually works: Git and repositories, issues and pull requests, Actions and Copilot, projects, permissions, and the open-source community around it.

Leading AI adoption rather than building it. The largest part is business value: what separates generative AI from prediction and classification, what pretrained models mean for how a project starts, what actually drives cost and what a return needs before it can be claimed, and where fabrications, unreliability and bias decide which use cases go first. Then the Microsoft services and what each is for: Copilot in the applications people already work in, Copilot Studio for a defined assistant, Foundry and Azure AI Search for what you build into a product, Microsoft Graph behind the personal answer, and the build, buy or extend decision. Finally the part most programmes get wrong: an AI council and what it decides, responsible AI as an assessment rather than a policy, adoption teams and champions, the barriers that leave licences unopened, and choosing between per-person and usage-based cost. Written for people who decide and sponsor rather than implement, so the questions are about choices and their consequences.

Administering Copilot and agents in Microsoft 365: which admin centre owns which object, how licensing decides what a user sees, the identity and conditional access features underneath, Purview information protection, data loss prevention, insider risk and lifecycle management, how Copilot inherits a user's existing permissions and what oversharing therefore means, and the practical work of assigning licences, monitoring adoption, creating agents and approving them.