Copilot Chat, Prompts and Agents

Microsoft via Coursera

Go to class Write review

Details

Go to class

Provider

Coursera
Pricing

Paid Course
Languages

English
Certificate

Certificate Available
Effort

10 weeks, 1 hour a week
Sessions

Self-Paced
Level

Beginner
Subtitles

English

Found in

Overview

Google, IBM & Meta Certificates – 40% Off

One plan covers every Professional Certificate on Coursera.

Unlock All Certificates

Gain expertise in the art of iterative prompting and specialized AI agents. This course teaches you to act as an orchestrator, using core principles—clarity, context, and constraints—to generate professional-grade business outputs. You’ll learn to navigate the critical divide between Web and Work data and move from general chat into specialized workflows using the Researcher and Analyst agents within Microsoft 365 Copilot. By mastering the prompt-and-verify loop, you’ll reduce back-and-forth and turn Copilot into a high-precision business partner. Through hands-on practice, you’ll learn how to select the right data source for different tasks, refine prompts through targeted follow-up questions, and improve the quality, relevance, and accuracy of AI-generated outputs. You’ll also explore how specialized agents support research, data interrogation, and business analysis while maintaining control over context and intent. By the end of the course, you’ll be able to structure effective prompts, orchestrate agent workflows, evaluate AI outputs for validity, and apply practical safeguards that protect sensitive information and support responsible business use.

Syllabus

Cut Your Prep Time in Half: The Efficiency of Web and Work Mode

This module establishes the foundational distinction every Copilot Chat user needs to understand: the difference between grounding AI responses in the public internet versus your organization's private data. Learners will leave with a clear framework for mode selection and the confidence to apply it consistently across their daily work.

Choosing Between Chat and Agents in Your Work

Copilot Chat is powerful on its own, but some tasks require more than a conversation. This module introduces the distinction between general chat and specialized agent-based experiences, giving learners the analytical framework to identify when a task warrants an agent and how to recognize the difference in practice.

Connecting Copilot to Your Full Data Ecosystem

Microsoft Graph is not limited to Microsoft 365 data. This module introduces Graph connectors, the mechanism that allows Copilot to draw from external organizational data sources like Salesforce, ServiceNow, or internal databases, giving learners a complete picture of how Copilot's Work mode can be extended across the full data ecosystem of a modern organization.

Ending the Rework Loop: Using Structure to Drive AI Quality

This module provides learners with a structured, repeatable method for writing prompts that yield professional-grade outputs on the first attempt. The Clarity, Context, and Constraints framework replaces guesswork with intention and reframes technical-skill prompts as something every professional already knows how to do: giving clear instructions to a capable colleague.

Precision Prompting: The Iterative Loop

Even a well-structured prompt may not yield a finished product on the first attempt. This module teaches learners to treat Copilot as a collaborative drafting partner, using targeted follow-up prompts to refine tone, restructure content, and verify factual accuracy through a deliberate iterative process. Think of it as giving feedback to a capable intern: not starting over, but redirecting with precision.

Orchestrating Agents: The Agent Landscape—Store vs. Custom

Before learners can orchestrate agents effectively, they need to understand the landscape. This module introduces the two paths to agent capability—the Agent Store for pre-built solutions and Microsoft Copilot Studio for custom builds—and provides learners with a clear framework for deciding which path best fits a given business need.

Orchestrating Agents: Researcher and Analyst Agent Execution

This module moves from understanding agents to directing them. Learners will apply a task analysis framework to select between the Researcher and Analyst agents, then practice invoking and grounding agents using the @mention syntax across real business scenarios, including deep market research and complex multi-source data analysis.

Orchestrating Agents: Configuring Custom Agents in Microsoft Copilot Studio

Pre-built agents cover a lot of ground, but some business needs require a custom solution. This module introduces Microsoft Copilot Studio as a platform for building agents tailored to specific organizational workflows, giving learners enough understanding of the configuration process to evaluate feasibility, brief a technical team, or build a basic agent themselves.

GenAI Module: Responsible Prompting

Writing effective prompts is a skill; writing responsible prompts is a professional standard. This module extends the Clarity, Context, and Constraints framework into the territory of data privacy and intellectual property, equipping learners with practical techniques for producing the outputs they need without exposing sensitive information or creating IP risk in client-facing work.

Project Module : The Helix Health AI Standard: Prompt Library & Agent Architecture Brief

Learners step into the role of a Lead Analyst at Helix Health, a diagnostic hardware company that went all-in on Microsoft 365's Microsoft Copilot six months ago. The tool is capable. The problem is that the team has not yet developed the skills or shared standards to use it effectively. Without a common framework for mode selection, grounding, and prompt structure, every team member is experimenting independently and getting inconsistent results. The shared #AI-Experimentation channel has become a collection of fragmented, underspecified prompts that are not giving Microsoft Copilot enough to work with. Your job is to close that gap. You will audit the team's Fragmented Files, apply the Clarity, Context, and Constraints framework to repair each prompt into a professional Enterprise Prompt Library, and design an Agent Workflow that solves the P1 Diagnostic Sync bottleneck—a process where technicians are manually bridging Microsoft Teams transcripts, technical manuals, and ServiceNow maintenance logs every time a lab sequencer goes offline.