Lesson 2 of 5 · about 5 min
Cleaning messy data
When you finish, you can describe data problems to AI, get a clear cleaning plan and check that nothing important was lost.
Stuck? Ask the AI tutorWhat messy data looks like
Real data is rarely tidy. The same city is spelled three ways. Dates are typed as 3/4, 4 March and March 4th. Names have extra spaces. Numbers are stored as text. Some rows are duplicates. Any summary built on messy data will be wrong or confusing.
Cleaning is slow work, and it is a good job to share with AI. But the order matters: look first, plan second, change third. Skipping to “fix everything” is how good data gets damaged.
Look first, with a small sample
Paste or share a small sample, not the whole file, and remove private details first. Ask: “List the problems you can see in this sample. Do not change anything yet.” The tool might spot mixed date formats, blank cells, spelling variants and likely duplicates.
Now you have a checklist. Ask which problems matter for your goal. If you only want totals by month, date format matters a lot and nickname spelling matters little. Cleaning for a purpose saves time.
Ask for steps you can repeat
A weak request is “Clean this data.” A strong request is: “Column D holds customer names typed many ways. I want them in Title Case with no extra spaces. Give me step-by-step instructions or a formula for my spreadsheet program, with a before and after example.”
Steps and formulas are better than a pasted, cleaned copy for two reasons. You can repeat them next month, and you can see exactly what changed. If the AI rewrites your data directly, it may quietly alter values, and you will not know.
Check after cleaning
After cleaning, compare the new data with the old. Count rows before and after. If you removed duplicates, check a few that were removed. Make sure totals you know are still the same. Keep an untouched copy of the original file in case you need to go back.
Worked example: a volunteer group in Bogotá has a sign-up list from three sources. Names repeat, phone numbers have mixed formats. The AI proposes steps to trim spaces and standardize numbers, and a way to flag likely duplicates for human review instead of deleting them. The volunteer coordinator decides which duplicates are real.
Practice now
- Make a small messy sample of ten rows with made-up names, dates and amounts. Ask an AI tool to list the problems without changing anything, then ask for repeatable cleaning steps.
- Run the steps on your sample and compare row counts and one total before and after. Write down anything the cleaning changed that you did not expect.
Use any AI assistant you have. Do not paste private, client or patient information into it.
Check yourself
1. What is the best first request when you share messy data?
2. Why ask for steps or formulas rather than a cleaned copy?
3. What should you do before cleaning begins?
Ask the AI tutor
Ask anything about this lesson, in any language. The tutor is an AI: it can be wrong, and it never needs your personal details.
AI-drafted. Checked by an automated validator and a separate AI reviewer; a person has not read every lesson. Translations and tutor answers are AI-made. Found a mistake? Tell us. AI can be wrong. General information only, not professional advice. Not affiliated with any AI company named. Current University is not accredited.