AI Essentials with Microsoft Copilot

Copilot licenses don't deliver value on their own. When staff send vague requests, they get generic drafts, confident-sounding errors, and rework that cancels out the time saved. This course gives your team a shared, repeatable way to guide Copilot in Word, Excel, and PowerPoint.

Risk and quality controls: Learners build a reusable prompt toolkit:

• CRE (context, role, expectation) to structure requests
• Zero-, one-, and few-shot examples for style and format
• "Ask two" clarifying questions to reduce guessing
• A safety line that limits Copilot to the supplied document, has it reply "not found" when information is missing, and keeps PII out of the output

The course also covers where AI makes things up and how to use Copilot for practical creation tasks, including documents, spreadsheets, presentations, and infographics.

Role fit: This is beginner-level training for project managers, analysts, admins, and marketers. No data analysis background is required for the Excel skill. The course does not cover IT setup, licensing, admin settings, or Copilot Studio agents, so pair it with admin-focused training for your M365 administrators.

Time investment: About 5 hours per learner across six skills. A single business case runs through every skill, which makes it easy to schedule in weekly blocks.

Change management: A common prompt framework gives Team Leads one standard to reinforce after rollout. Use Playlists to sequence this course with follow-on AI training.

Most teams already have Copilot inside Word, Excel, and PowerPoint. The gap is knowing how to direct it. This course builds that skill through one connected business scenario: a fictional wellness startup, Thrive 365, carried from document to spreadsheet to slide deck. Your team practices the same reusable prompt workflow at each step.

• Faster, more consistent first drafts in Word: Learners move from loose, naive requests to structured prompts using CRE (context, role, expectation). They add examples only when they need them, then refine a full business document with service tiers, onboarding steps, and check-ins.
• Fewer AI errors reaching stakeholders: Teams add an "ask two" step, where Copilot asks two clarifying questions before it writes. They also add a safety line, such as "use only this document, reply 'not found' if info is missing, and leave out PII." Both habits reduce the risk of hallucinated or sensitive content.
• Pitch decks and budget asks built with clearer constraints: In PowerPoint, learners use structured prompts to shape presentation output around the audience, slide limits, and a clear call to action. The same approach helps teams prepare internal funding requests.
• More structured financial work in Excel: Learners apply the prompt workflow to Excel-based financial work, using clear context and expectations so spreadsheet output is easier to review, revise, and explain. Teams see how the same prompt engineering habits apply whether the spreadsheet task is simple or more involved.

The course also sets realistic expectations. Learners test Copilot Create on infographics and see where it falls short, so they know when AI output still needs a person to finish it.

Knowledge

• How classic rule-based software (Software 1.0) differs from machine learning, deep learning, and generative AI, and where Copilot fits in that mental model
• Why AI models can make things up while sounding confident, and how that risk shapes everyday Copilot use
• The CRE framework (context, role, expectation) for writing prompts in plain English
• When to use zero-shot, one-shot, or few-shot prompting, and why more than two examples usually adds noise
• The difference between excerpts (the facts you supply) and examples (the style and format you want)
• Known Copilot limits, including why outputs need review and where Copilot Create fits for image and infographic work

Ability

• Turn rough requests into structured Word prompts that produce complete, editable business documents, such as service packages, onboarding steps, and pricing tables
• Build and reorganize PowerPoint decks from a blank file or existing documents, with set section order, slide limits, and merged overlapping content
• Add an "ask two" clarifying-question step and a safety line that limits Copilot to supplied source material, returns "not found" when information is missing, and excludes PII
• Use Copilot in Excel to work with formulas and structured spreadsheets that update when inputs change
• Use Copilot Chat to write role-based image prompts and identify how Copilot Create supports generating and modifying images
• Reuse one prompt toolkit across Word, PowerPoint, Excel, and Copilot Chat so teams get consistent results