Getting Set Up for Free
Before you can point AI at your data, you need Power BI installed and a small practice dataset to play with. The good news: everything you need costs nothing. This lesson walks you through downloading Power BI Desktop, getting a free AI assistant open in a tab, and loading a sample dataset so you have real numbers to work with in the next module.
Take your time here. A clean setup now saves a lot of confusion later. By the end of this lesson you will have Power BI open, data loaded, and an AI assistant ready to help.
What You'll Learn
- How to download and install Power BI Desktop for free
- Which free AI assistant to keep open alongside Power BI
- How to get a practice dataset if you do not have your own
- How to load data into Power BI so it is ready for cleaning
Step 1: Get Power BI Desktop
Power BI Desktop is the free Windows app where you build reports. To get it, search "Power BI Desktop download" and get it from Microsoft, either from the Microsoft Store (easiest, it auto-updates) or as a direct installer from Microsoft's site. Install it like any other app and open it.
A note for Mac and Chromebook users: Power BI Desktop is Windows-only. If you are on a Mac or Chromebook, you have options: use the free Power BI web service at the Power BI site (sign in with a school or work email), run Windows in a virtual machine, or simply follow along conceptually and practice the AI-prompting skills, which transfer to any tool. Do not let the platform stop you; the AI skills are the point.
When Power BI opens, you will see a blank canvas with ribbons at the top. The three areas you will use most are Home (to get data), Power Query (to clean it, opened via "Transform data"), and the report canvas (to build visuals). Do not worry about memorizing this; you will learn each area as you need it.
Step 2: Open a Free AI Assistant
Keep one of these open in a browser tab next to Power BI:
- ChatGPT (chat.openai.com) is a strong all-rounder and great at DAX.
- Claude (claude.ai) writes clear explanations and clean formulas.
- Gemini (gemini.google.com) integrates with Google data and is solid at DAX.
- Perplexity (perplexity.ai) is best when you want cited, up-to-date answers.
Any of these works for this course. If you are unsure, start with ChatGPT or Claude. All have free tiers that are more than enough. You will paste your table names and questions here whenever you need DAX, a cleaning plan, or an explanation.
Test it works with this prompt:
"I'm about to learn Power BI. In three bullet points, tell me what to double-check every time I load a new dataset."
If you get a sensible answer, you are ready.
Step 3: Get a Practice Dataset
You need some data to practice on. Three easy options:
Option A: Use a built-in sample. Search online for the "Power BI financial sample" Excel file, a small, free spreadsheet Microsoft provides with sales, segments, countries, and dates. It is perfect for beginners.
Option B: Ask AI to generate one. Paste this into your assistant:
"Generate a CSV of 40 rows of fake sales data with these columns: OrderDate, Region, Product, Category, Quantity, UnitPrice, Revenue. Include a few messy values on purpose: some blank cells, inconsistent region names like 'USA' and 'United States', and a couple of dates written as text. Output it as plain CSV I can paste into a file."
Copy the result into a text file, save it as sales.csv, and you have a realistic messy dataset, ideal for practicing the cleaning skills coming next.
Option C: Use your own data. A club budget, a personal expense export, a spreadsheet from an internship. Real data you care about makes learning stick. Just remove anything sensitive first.
Step 4: Load Data Into Power BI
With Power BI Desktop open:
- On the Home ribbon, click Get data.
- Choose your source: Excel workbook for
.xlsx, or Text/CSV for.csv. - Browse to your file and select it.
- In the preview window, click Transform data (not "Load"). This opens Power Query, where you will clean the data in the next lesson.
Clicking "Transform data" instead of "Load" is a good habit: it takes you straight to the cleaning stage rather than dumping raw, messy data into your model. If you accidentally clicked "Load," no problem, you can open Power Query anytime with the Transform data button on the Home ribbon.
A Quick Sanity Check
Once your data is in Power Query, glance at it and ask yourself three questions (or ask your AI assistant):
- Do the column headers look right, or did the first row of data become the headers?
- Are numbers stored as numbers and dates as dates? (Look at the little icon in each column header.)
- Are there obvious blanks, typos, or inconsistent labels?
You do not need to fix anything yet. Just noticing these issues prepares you for the cleaning lesson. If you want, paste your column list into your assistant and ask:
"Here are my Power BI columns and a few sample rows: [paste]. What data-quality problems should I watch for before I start building charts?"
The assistant will point out likely issues, which is a preview of the AI-assisted cleaning you will do next.
Key Takeaways
- Power BI Desktop is a free Windows app; Mac and Chromebook users can use the free web service or focus on the transferable AI skills.
- Keep a free AI assistant (ChatGPT, Claude, Gemini, or Perplexity) open in a tab to help with DAX, cleaning, and explanations.
- Get practice data from the Microsoft financial sample, by asking AI to generate a messy CSV, or from your own de-sensitized spreadsheet.
- Load data with Get data, then click Transform data to go straight into Power Query for cleaning.
- Do a quick sanity check on headers, data types, and obvious messiness before building anything.