
Before we build anything, let's get you fluent in the new Copilot pane. It has changed more in 2026 than in the previous two years together: there are now three ways of working, a choice of AI models, built-in skills, and a review step after every change.
Practice files (Resources): the full student pack (all datasets, both custom skills and the Prompt Playbook), the Prompt Playbook PDF and 01_Contoso-Coffee_SalesData.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Big jobs deserve a plan. In this lecture you'll ask for a whole monthly report in a single prompt, let Copilot plan it first, change the plan just by chatting, and only then let it build. Then we'll switch to a reasoning model for a hard why question.
Practice files (Resources): 01_Contoso-Coffee_SalesData.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
A default PivotTable and a default chart are fine for you. They are not fine for your boss. In this lecture we build the analysis with two prompts, and then use two of Copilot's design skills to make it boardroom-ready.
Practice files (Resources): 01_Contoso-Coffee_SalesData.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
If you keep typing the same instructions, no merged cells, dollars, name my tables properly, stop. Copilot has two places to store your standards: personalisation, which follows you into every workbook, and workbook rules, which live inside one file and apply to anyone who uses it.
Practice files (Resources): 01_Contoso-Coffee_SalesData.xlsx and 11_Workbook-Rules-Starter.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Copilot only knows what's in your workbook, unless you give it more. In this lecture we attach a pricing memo from finance, have Copilot apply it to a pricing model, check where every number came from, and then challenge the result.
Practice files (Resources): 05_Pricing-and-Demand_Model.xlsx and 10_Contoso-Coffee_Q4-Pricing-Memo.docx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Python in Excel is now something you can ask for in plain English. In this lecture you'll see two different ways Copilot uses Python: as an analysis that gives you an answer, and as live Python cells that stay in your workbook and recalculate.
Practice files (Resources): 01_Contoso-Coffee_SalesData.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Copilot can change hundreds of cells in seconds. So you need to know exactly what changed, who changed it, and how to undo just the AI part. In this lecture we use change history to answer all three.
Practice files (Resources): 12_H2-Forecast-Model_Change-History.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Sometimes you don't have a workbook yet, just an idea. The Excel agent in the Microsoft 365 Copilot app builds a complete, working workbook from a description. Let's build a cash-flow forecast from one paragraph, and then improve it just by chatting.
No practice file needed: this lecture starts from a blank page in the Microsoft 365 Copilot app.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Every analyst knows this file: a messy export with dates in three formats, regions spelled four different ways, and duplicates. In this lecture we clean it the professional way: scan first, fix on a copy, prove the risky decisions with evidence, and reconcile at the end.
Practice files (Resources): 02_Contoso-Coffee_SalesData-MESSY.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
If you do the same review every month, you shouldn't explain it to Copilot every month. In this lecture you teach Copilot your process once, as a custom skill: a small text file in your OneDrive. After that, the whole review is one line.
Practice files (Resources): 13_Q3-Opex-Close.xlsx and the month-end-variance-review custom skill (zip with SKILL.md).
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Here's a situation you'll recognise. A colleague leaves, and you inherit their investment model. The board meets on Monday, and the model says go. In this lecture we audit it with Copilot's check-model skill, and then we verify what it finds, because an audit you don't check is just another opinion.
Practice files (Resources): 16_Inherited-Model_Rescue.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
In part one, check-model found seven errors and turned our go into a no-go. Now we fix it like a professional: on a copy, with a change log. Then we prove every fix with an NPV bridge, find the break-even, and write the message to the board.
Practice files (Resources): 16_Inherited-Model_Rescue.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Every Monday, someone in your team loses a morning to this: two sales exports that never match, and a list of store targets. In this two-part case study we automate the whole job. Part one: clean and combine, with rules you write like a policy, and a reconciliation to the cent.
Practice files (Resources): 14c_Weekly-KPI_W41_with-Targets.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
In part one we built a clean, reconciled sales table. Now the payoff: the Monday report itself, designed for executives. And then the real goal: making next Monday a one-line job.
Practice files (Resources): 14d_Weekly-KPI_W42_with-Targets.xlsx and the weekly-kpi-pack custom skill (zip with SKILL.md).
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
400 customer comments, and the board wants one slide by Friday. Reading them all takes a day. Asking AI for a quick summary takes a minute, but can you trust it? In this lecture we turn comments into labels you can count, check and defend.
Practice files (Resources): 17_Customer-Feedback_400.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
In part one we learned what customers complain about. Now the questions a board actually asks: is it one store or everywhere? When did the app problem start? And what can we claim, and what still needs checking?
Practice files (Resources): 17_Customer-Feedback_400.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
Copilot changes every month: new models, new modes, new skills. So how do you know it still gets your numbers right? You test it, like any system you rely on. In this lecture you build a small evaluation kit: fixed questions with known answers, run across modes and models, and scored yes or no.
Practice files (Resources): 18_Copilot-Eval-Kit.xlsx and 01_Contoso-Coffee_SalesData.xlsx.
Copilot changes often, so your screen may differ slightly from the video. Where it matters, features are labeled as available now, rolling out or preview.
This course contains the use of artificial intelligence.
Stop memorizing Excel formulas. Start telling Excel what you want — and let AI do the heavy lifting.
Microsoft Copilot has changed what it means to be good at Excel. The professionals who win in the AI-powered workplace aren’t the ones who remember every function — they’re the ones who can ask the right questions and turn raw data into decisions in minutes. This course teaches you exactly how.
This is a practical, hands-on masterclass built around real business scenarios, not theory. From your very first prompt to a full capstone project, you’ll work inside Excel with Copilot the way you actually would on the job — analyzing sales, cleaning messy data, building dashboards, forecasting, and automating the tasks that used to eat your afternoons.
What you’ll learn:
Set up Copilot in Excel and write structured, reliable prompts using a simple, repeatable framework
Summarize and analyze data in seconds — and ask follow-up questions like a conversation
Generate, explain and fix complex formulas without memorizing syntax
Clean, reshape and standardize messy spreadsheets fast
Build charts, dashboards and decision-ready reports your stakeholders will trust
Forecast, run what-if scenarios, and support real decisions with AI insight
Extend Excel into PowerPoint, Word and Outlook for complete end-to-end workflows
Automate repetitive, multi-step tasks and reclaim hours every week
Turn numbers into a persuasive data story
Apply advanced “power prompt” patterns that separate everyday users from true experts
Use Copilot responsibly — knowing its limits, verifying output, and applying best-practice governance and ethics
Who this is for: Analysts, finance and marketing professionals, operations teams, managers, and anyone who lives in Excel and wants to work faster and smarter with AI. No coding required.
How you’ll learn: Every concept is taught by doing. You’ll download the same datasets I use on screen, follow along step by step, complete practice exercises, and finish with a capstone project — using Copilot in Excel to drive a regional expansion decision from raw data to a confident recommendation.
By the end, you won’t just know what Copilot can do — you’ll have the practical, job-ready skills to get measurably more done every single day.
About your instructor: I am Alaaldin Ahmed, a Senior AI Workforce Specialist at Microsoft. I spent years doing this work the classic way — through my MBA and across enterprise data projects — long before generative AI existed. I built this course to teach the exact shift I made myself. Views expressed are my own and do not represent Microsoft.
You get full lifetime access, every dataset and exercise file I use on screen, and Udemy’s 30-day money-back guarantee — so there is no risk in starting today.
Enroll now and start working the AI-powered way.