ExamFinch.com My Study

Study guide

9.5 Adding Knowledge Sources to In-App Help & Guidance in Dynamics 365 Finance & Supply Chain

Practice this exam

Microsoft Certified: Agentic AI Business Solutions Architect Expert AB-100

Track your progress as you read

Finished sections get a check mark.

1 Chapter 1: Requirements Analysis & Grounding Data

1.1 Assessing Agents for Task Automation, Analytics & Decision-Making 1.2 Grounding Data Evaluation: Accuracy, Relevance, Timeliness, Cleanliness & Availability 1.3 Structuring Enterprise Business Data for AI System Interoperability

2 Chapter 2: AI Strategy, Cloud Adoption Framework & CoE

2.1 Cloud Adoption Framework (CAF) for Azure AI Strategy & Roadmap 2.2 Multi-Agent Architecture: Microsoft 365 Copilot, Copilot Studio & Foundry 2.3 Microsoft AI Center of Excellence (CoE) & Solution Governance 2.4 Developing Use Cases for Prebuilt Agents & the Microsoft Agent Catalog 2.5 Solution Rules & Constraints for Copilot Studio, Microsoft Foundry & Foundry Tools

3 Chapter 3: Model Selection, Prompt Engineering & Dynamics AI

3.1 Model Selection Strategy, Custom Models & Small Language Models (SLMs) 3.2 Enterprise Prompt Engineering Guidelines & Prompt Libraries 3.3 Cross-App AI Strategy for Multi-App Dynamics 365 Solutions

4 Chapter 4: ROI, TCO, Build vs Buy & Model Routing

4.1 ROI Criteria & Total Cost of Ownership (TCO) for AI Solutions 4.2 Comprehensive Financial ROI Analysis for Business Processes 4.3 Build vs Buy vs Extend Architectural Decision Framework 4.4 Dynamic Model Router Architecture & Cost-Performance Optimization

5 Chapter 5: Dynamics 365 AI & Contact Center Design

5.1 Business Terms & Copilot Customization in Customer Experience and Service 5.2 Connectors & Data Flow Design for Copilot in Dynamics 365 Sales 5.3 Agent Integration with Dynamics 365 Contact Center Channels

6 Chapter 6: Agent Archetypes & Copilot Studio Core

6.1 Designing Task Agents, Autonomous Agents & Prompt-and-Response Agents 6.2 Copilot Studio Topic Design, Trigger Phrases & Graceful Fallbacks 6.3 Orchestration Selection: Standard NLP vs CLU vs Generative AI Orchestration

7 Chapter 7: Copilot Studio Advanced Actions & Power Platform

7.1 Designing Agent Flows & Prompt Actions in Copilot Studio 7.2 Power Apps Canvas App Integration, Generative Pages & Agent Feeds 7.3 Power Platform Well-Architected Framework for AI Workloads 7.4 Proposing Microsoft Power Platform AI Features & the AI Hub

8 Chapter 8: Agent Extensibility, MCP & Computer Use

8.1 Microsoft Foundry Custom Models & Foundry Tools Selection 8.2 Agent Extensibility via Model Context Protocol (MCP) in Copilot Studio 8.3 Computer Use Automation & Web UI Interactions in Copilot Studio

9 Chapter 9: Advanced Behaviors & Prebuilt App Orchestration

9.1 Agent Behaviors: Multi-Step Reasoning, Voice Mode & M365 Agent Optimization 9.2 Orchestrating Prebuilt AI in Dynamics 365 F&O and M365 Copilot for Sales/Service 9.3 Orchestrating Prebuilt Copilot AI Features in Dynamics 365 Customer Experience & Service 9.4 Proposing Microsoft 365 Agents for Business Scenarios 9.5 Adding Knowledge Sources to In-App Help & Guidance in Dynamics 365 Finance & Supply Chain

10 Chapter 10: Telemetry, Monitoring & Agent Tuning

10.1 Monitoring Architectures: Azure Monitor, Application Insights & Studio Analytics 10.2 Telemetry Data Interpretation, Performance Metrics & Latency Tuning 10.3 User Feedback Analysis, Backlog Prioritization & Continuous Improvement

11 Chapter 11: Testing Strategy & Model Validation

11.1 Agent Testing Frameworks, Quality Metrics & Groundedness Benchmarks 11.2 Custom AI Model Validation, Safety Criteria & Regression Testing 11.3 Copilot-Assisted Test Case Generation & Cross-Dynamics End-to-End Testing 11.4 Validating Effective Copilot Prompt Practices

12 Chapter 12: ALM for Grounding Data & Copilot Studio Agents

12.1 ALM for Grounding Data, Vector Indexes & Knowledge Synchronization 12.2 ALM for Copilot Studio Agents, Connectors & Solution Packaging 12.3 Environment Strategy & Automated Deployment Pipelines for Studio Agents

13 Chapter 13: ALM for Foundry Agents & Dynamics 365 AI

13.1 ALM for Microsoft Foundry Agents Service & Model Registries 13.2 ALM for Custom AI Models, Fine-Tuning Artifacts & Deployment Slots 13.3 ALM for Dynamics 365 AI in Finance, Supply Chain & Customer Engagement

14 Chapter 14: Agent Identity, Security & Access Controls

14.1 Agent Identity Architecture, Microsoft Entra ID & Authentication Flows 14.2 Access Controls, Role-Based Security & Permissions on Grounding Data 14.3 Secure Data Processing & Grounding Data Isolation

15 Chapter 15: Model Security & Threat Mitigation

15.1 Model Security Architecture & Defense Against Vulnerabilities 15.2 Mitigating Prompt Manipulation, Jailbreaks & Indirect Prompt Injections 15.3 Azure AI Content Safety, Prompt Shields & Guardrails

16 Chapter 16: Responsible AI, Governance, Residency & Audit

16.1 Microsoft Responsible AI Standard: Principles, Assessment & Implementation 16.2 Data Residency, Sovereign Boundaries & Cross-Border Compliance 16.3 Comprehensive Audit Trails, Model Decision Logging & Microsoft Purview 16.4 Designing Governance for Agents: Environment Strategy, Data Policies & Agent Inventory

9.5 Adding Knowledge Sources to In-App Help & Guidance in Dynamics 365 Finance & Supply Chain

Key Takeaways Explain This Topic AI

Generative help and guidance with Copilot is grounded by default in public Microsoft Learn documentation for finance and operations apps, using a query rewrite, a Bing index search over learn.microsoft.com, and a large language model response with citations.

Custom knowledge is added by opening Copilot Studio in the Dataverse environment associated with the finance and operations environment, opening the agent named Copilot for finance and operations apps, adding knowledge on the Knowledge tab, waiting for Status to reach Ready, testing, and publishing.

Uploaded files such as PDF, RTF and Word are supported directly, but SharePoint and other data sources require adding a custom topic to the Copilot for finance and operations apps agent rather than relying on the out-of-box conversational boosting topic.

Dataverse virtual entities published from finance and operations apps are not currently supported as knowledge sources, and AI general knowledge is disabled by default and consulted only after custom knowledge sources are exhausted.

The classic Help pane is extended separately by publishing a custom Help website with title, description, ms.locale, ms.search.form and ms.search.scope metadata, indexing it with ConvertHtmlToJson into an Azure search service, and connecting it through the Visual Studio HelppaneMacro and Language.config files.

Last updated: September 2026

Adding Knowledge Sources to In-App Help & Guidance in Dynamics 365 Finance & Supply Chain

Quick Answer: Dynamics 365 finance and operations apps expose two distinct in-app help systems , and the AB-100 blueprint expects an architect to recommend the right process for each. The classic Help pane (opened with the ? button or Ctrl+Shift+? ) serves Microsoft Learn documentation plus task guides , and is extended by publishing a custom Help website , indexing it with an Azure search service , and connecting both through a Visual Studio extension whose HelppaneMacro file carries the web app name, search service name, and query API key, with Language.config mapping indexes to locales. The newer generative help and guidance with Copilot is a conversational sidecar grounded by default in public Microsoft Learn documentation only , and is extended in Copilot Studio by opening the agent named Copilot for finance and operations apps , adding knowledge on the Knowledge tab, waiting for Status = Ready , testing, and publishing . Two hard constraints define the design: Dataverse virtual entities published from finance and operations apps are not supported as knowledge sources , and AI general knowledge (the language model's own knowledge plus Bing-identified web content) is disabled by default and is only consulted after custom knowledge sources are exhausted.

This is one of the most frequently mis-answered areas of the AB-100 blueprint, because architects assume the two help systems share a knowledge store. They do not. Uploading a PDF to Copilot Studio does not put it in the Help pane. Publishing a custom Help website does not make Copilot cite it. Recommending the wrong process produces a project that delivers help content users never see.

1. Two Help Systems, Two Extension Processes

  • Dimension Help pane (classic) Generative help and guidance with Copilot
  • Entry point ? button > Help , or Ctrl+Shift+? Copilot button at the top of the page (sidecar)

Default content Microsoft Learn documentation, context-sensitive to the current page Public Microsoft Learn documentation for finance and operations apps

Additional content type Task guides (interactive, step-by-step, playable) Conversational generated answers with citations

Extension mechanism Custom Help website + Azure search index + Visual Studio Help-pane extension Knowledge sources added to a Copilot Studio agent

Who performs it Developer with finance and operations development tools in Visual Studio Administrator or maker in Copilot Studio

Applies to Finance, Supply Chain Management, Commerce, Human Resources Finance and operations apps with Copilot enabled

Both systems are available to users of Dynamics 365 Commerce, Finance, Human Resources, and Supply Chain Management .

  • 2. Extending Generative Help and Guidance with Copilot
  • 2.1 How It Works

Copilot provides in-app help and guidance using generative AI to give contextual support. Its response workflow has three steps:

Query rewrite — Copilot reformulates the user's question to optimise understanding and search efficacy.

Document search — Copilot uses the Bing search index to find the most relevant documents in the learn.microsoft.com domain.

Response generation — a large language model constructs an answer with citations from the source material.

By default, Copilot knowledge is limited to the published product documentation for finance and operations apps, and it respects each user's selected language preferences.

Underneath, the capability uses the out-of-box conversational boosting topic, which relies on generative answers in Copilot Studio to answer from the knowledge sources attached to it.

2.2 The Process for Adding Custom Knowledge

To add specific knowledge:

Open Copilot Studio and select the Dataverse environment associated with your finance and operations apps environment . (Choosing the wrong environment is the single most common failure.)

On the left navigation pane, select Agents .

Open the agent named Copilot for finance and operations apps .

On the Knowledge tab, select Add knowledge , then choose the knowledge-source type. Use Upload files to add files from the local computer — supported formats include PDF, RTF, and Word .

The source is added to the grid. Track processing in the Status column; availability takes a few minutes.

When Status shows Ready , select Test on the toolbar and enter test questions to confirm the source behaves as expected.

Select Publish to roll the change out to users in the finance and operations apps.

For knowledge types beyond uploaded files — SharePoint or other data sources — the architect must add a custom topic to the Copilot for finance and operations apps agent in Copilot Studio rather than relying on the out-of-box conversational boosting topic alone.

Tip that saves a support ticket: to make new knowledge available after publishing, users must close any open Copilot sessions and start a new session .

2.3 Enabling AI General Knowledge

General knowledge lets Copilot answer general questions using content that is part of the language model, web content identified through Bing Search , and other sources. Copilot uses this only after it exhausts your custom knowledge sources . It is disabled by default :

Open Copilot Studio and select the Dataverse environment associated with the finance and operations environment.

Select Copilots in the navigation pane.

Open Copilot for finance and operations apps .

On the Overview tab, in the Knowledge section, set Allow the AI to use its own general knowledge to Enabled .

Select Test , enter test questions, then Publish .

Microsoft's own warning is the architectural point: enabling general knowledge increases the richness of responses and increases the risk that answers include incorrect content . It must be carefully considered and tested before publishing to production. For a regulated finance process, the defensible default is to leave it disabled and rely on curated sources.

2.4 Two Constraints to Design Around

Dataverse virtual entities published from finance and operations apps are not currently supported as knowledge sources. An architect who plans to ground the help agent on ERP master data through virtual entities must choose a different pattern.

The conversational boosting topic assumes all knowledge relates to product functionality. Adding unrelated knowledge can make Copilot imply that the added content describes how to use finance and operations apps. Knowledge curation is therefore a correctness control, not housekeeping.

The amount of content that can be added is limited by knowledge-source type , following Copilot Studio's published knowledge limits.

2.5 Feedback as a Quality Loop

Users rate responses with thumbs-up and thumbs-down in the sidecar. Turning on enhanced user feedback for Copilot adds written feedback. This feedback is the measurable input to the continuous-improvement backlog for help content — an architect should specify it as part of the deployment, not leave it to chance.

3. Extending the Classic Help Pane with a Custom Help Website

This is a developer process, performed once, using the finance and operations development tools in Visual Studio. Custom Help content typically originates from Microsoft documentation repositories, converted Dynamics AX content, or purpose-written HTML.

3.1 The Four-Step Process

Create the custom Help content.

Publish it on a website — existing or dedicated, public or private, but users must not be required to sign in .

Index the content with a search service.

Connect the Help pane to the website and the search service.

3.2 Required Metadata

For context-sensitive help and full-text search to return results, each article must carry:

  • Property Purpose
  • title Value for full-text search from the Help pane
  • description Value for full-text search from the Help pane
  • ms.search.form The Application Object Tree (AOT) name of a page — this is what makes help context-sensitive
  • ms.locale Article language, matched against browser locale
  • ms.search.scope Which client the article appears in: Core , Operations , Retail , Human Resources

ms.search.scope deserves emphasis: without the Core value the article does not appear in the Help pane at all . Operations targets Finance and Supply Chain Management solutions, Retail targets Commerce, and Human Resources targets Dynamics 365 Human Resources.

3.3 Indexing and Connecting

The ConvertHtmlToJson tool transforms HTML files into JSON files that are loaded into an Azure search service, which then generates context-sensitive links: ConvertHtmlToJson.exe --h <html path> -j <json path> --v <true|false>

The Help pane is then extended in Visual Studio by opening the AzureSearchCustomHelp.sln solution, importing the HelppaneOption.axpp project, and updating the HelppaneMacro file with the web app name , the search service admin/query API key , and the search service name . Index-to-language mapping is declared in Language.config using index="" , parentindex="" , and ultimateindex="" , which also defines language fallback order: the client language first, then the language named by parentlanguage , then the language set by ultimateindex . The build output is a model uploaded to the Asset library of the customer or solution project in Microsoft Dynamics Lifecycle Services (LCS) .

3.4 Task Guides

Task guides are controlled, guided, interactive walkthroughs played from the Help pane's Task guides tab. Before users can access them, a system administrator must configure settings on the System parameters page in Finance, Supply Chain Management, or Commerce. Organisations create their own task guides with Task recorder . Localised task guides ship in libraries whose title contains "All languages", and each user selects a task-guide language under Options > Preferences .

One safety warning that belongs in any deployment plan: data entered while playing a task guide is real . In a production environment it is written to the current company. Task-guide training must therefore be scheduled against a sandbox, not production.

Two scope exceptions: the Help-topics behaviour does not apply to Dynamics 365 Human Resources, whose Help system is automatically connected to task guides and for which custom task guides cannot be created .

4. Recommending the Right Process

Requirement Recommendation

Conversational answers over internal SOP documents inside the app Add file knowledge to Copilot for finance and operations apps in Copilot Studio, test, publish

Conversational answers over SharePoint or another data source Add a custom topic to that same agent — file upload alone is insufficient

Context-sensitive written help for customised forms Custom Help website with ms.search.form set to the AOT page name, indexed and connected via the Visual Studio extension

Guided, interactive walkthroughs of a customised process Task recorder task guides, with System parameters configured, played in a sandbox

Allow answers beyond curated content AI general knowledge — enabled deliberately, tested, and generally left off for regulated processes

Ground the help agent on ERP master data via virtual entities Not supported — redesign the grounding approach

AB-100 Exam Tip: Watch for the word conversational . If the scenario describes users asking questions in natural language and receiving cited answers, the process runs through Copilot Studio on the Copilot for finance and operations apps agent. If it describes context-sensitive articles for customised forms or step-by-step guided walkthroughs, the process runs through the custom Help website and task guides. Options that propose uploading PDFs to Copilot Studio to fix context-sensitive Help-pane links — or that propose a custom Help website to give Copilot new knowledge — are crossing the two systems and are wrong.

Loading diagram... Two Help Systems, Two Extension Processes: Copilot Generative Help versus the Classic Help Pane