Manage prompts and conversations by using AI
Free Microsoft Certified: AI Business Professional practice — 6 questions on Manage prompts and conversations by using AI, with explanations. No sign-up. Full 12-question mixed test →
A sales manager in Copilot Chat asks, "Summarize our Q3 performance," and receives a generic, unhelpful response because Copilot doesn't know what documents or metrics to draw from. The manager wants a better answer in the same session. What is the BEST next action?
Iterating in a multi-turn conversation by adding missing goal, context, expectations, and source details is the core technique for fixing a weak prompt and is directly tested as part of prompt refinement.
A finance team member has a large Excel workbook with several years of numeric sales data and wants Copilot to identify trends, build calculations, and generate a chart based on that data. Which capability is BEST suited for this task?
The Analyst reasoning agent is specifically built to work with numerical/tabular data, perform calculations via code execution, and surface trends and visualizations — exactly the scenario described.
An employee needs Copilot Chat to summarize one exact quarterly report file, not any similar-sounding documents that may exist elsewhere in the tenant. What is the BEST way to ensure Copilot grounds its answer strictly in that one document?
Using the "/" command to explicitly reference a specific file grounds Copilot's response in that exact source, which is the correct approach when precision matters more than open-ended search.
A team lead wants a custom agent tailored to their department's unique intake workflow, but no existing agent in the Agent Store meets this need. As a business user with no coding background, what is the appropriate way to create this agent?
Copilot Studio offers a consumer-level, natural-language, no-code way for business users to describe instructions and knowledge sources and build a custom agent without needing a developer.
A colleague argues that there is no real difference between "having a conversation with Copilot" and "using an agent," since both just respond to text input. Which statement BEST corrects this misunderstanding?
The key distinction tested is that agents are purpose-built and configured with instructions, grounding sources, and/or actions for repeatable use, unlike an unconfigured, one-time chat exchange.
A department wants to standardize how employees ask Copilot to draft weekly status reports across Word and Outlook so that results are consistent regardless of who writes the prompt. What is the BEST way to build this into a reusable prompt library?
A proper prompt library documents the reusable prompt structure — goal, context, expectations, and source — so teams can consistently apply and adapt proven prompts across multiple Microsoft 365 apps.
Ready for the real thing?
The full course: two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed explanations.
$99.99 $34.99 with code FREETEST33 — valid through September 19.
Get my $34.99 deal →