AB-210 Practice Questions

The issue is that the sales manager already generated a Sales Research Agent blueprint but forgot to include the quarterly competitive intelligence brief. The brief is an external document and the case study states it is delivered as a PDF document , so the correct first content action is to upload the competitive intelligence brief as a PDF file , not CSV. Microsoft’s Sales Research Agent documentation states that users can select Add data to upload files including CSV, Excel, or PDF to augment research with additional data.

The second required action is to add a data source , because the competitive intelligence brief is not part of the default Dynamics 365 Sales data. Sales Research Agent connects to Dynamics 365 Sales data by default, but additional external files must be added as supporting data sources before they can influence the analysis. The documentation identifies the Data area as the place where the research data source is specified and where additional data sources can be added.

The final action is to regenerate the canvas/blueprint so the agent reprocesses the analysis using the newly added PD

F.  

References/topics: Sales Research Agent; research canvas; adding PDF data sources; blueprint regeneration; intelligent opportunity research.

Requirement

Configuration

Blueprints

Name the blueprint workspace and pin the workspace.

Questions

Create a business function with seeded prompts.

For Blueprints, the correct configuration is to name the blueprint workspace and pin the workspace. The requirement is not asking to regenerate analysis; it asks for a reliable way to return to frequently used research output. In Sales Research Agent, a generated blueprint exists inside a workspace, and Microsoft states that the workspace displays the workspace name, can be renamed, and can be switched from the workspace menu. That makes naming and pinning the workspace the correct operational pattern for reusable research access.

For Questions, the correct configuration is to create a business function with seeded prompts. Sales Research Agent business functions are designed to tailor research to a specific use case, and administrators can add starter prompts. Microsoft explains that starter prompts appear on the Sales Research Agent home page for quick access and that the full prompt is sent to the agent when selected. This directly satisfies the requirement to retain a reusable library of approved questions.

Do not select regenerate the blueprint because regeneration refreshes or recreates analysis, but it does not create a persistent shortcut to frequently used research. Do not select custom table or manual prompt questions in Sales Copilot because the native Sales Research Agent feature already supports reusable business-function prompts.

References/topics: Sales Research Agent; research blueprints; workspace management; business functions; seeded/starter prompts; intelligent opportunity research.

Requirement

Configuration

Automatically prioritize leads based on conversion probability

Predictive lead scoring

Ensure the lead scoring improvement requirement is met

Automatically retrain model

The correct capability for automatically prioritizing leads based on conversion probability is Predictive lead scoring . Dynamics 365 Sales predictive lead scoring assigns scores to leads so sellers can prioritize the leads most likely to qualify or convert. Microsoft’s Sales documentation states that predictive lead scoring helps sales teams prioritize leads based on scores and improve lead qualification outcomes. This directly matches Contoso’s requirement to guide sellers toward high-potential leads and reduce inconsistent manual prioritization across regions.

To meet the requirement that lead scoring must continuously improve based on historical sales outcomes, configure the model to automatically retrain . Microsoft’s predictive lead scoring guidance explains that retraining uses the latest leads in the organization to improve model accuracy, and automatic retraining allows the application to retrain the model every 15 days. This is the correct option because Contoso wants the scoring model to improve continuously as more closed lead outcome data becomes available.

“Lead scoring model” is too generic; the AI capability required is specifically predictive lead scoring. “Relationship intelligence” supports relationship analytics and engagement insights, but it does not score leads by conversion probability. “Configure forecast” applies to pipeline forecasting, not lead scoring accuracy.

References/topics: Predictive lead scoring; lead prioritization; scoring model accuracy; automatic model retraining; AI-driven lead qualification.