How to Analyse a CSV File in Three Steps
You have a CSV. You want to know what is in it — how many rows, which values repeat, what the totals look like. Here is the whole process, with nothing to install and no account to create.
Step 1 — Drop the file in
Open the tool and drag your CSV anywhere onto the File Select panel. You can also click the panel to browse.
The file becomes a SQL table straight away. Its name is a single letter shown in the Alias column, taken from the start of the filename — sales.csv becomes S. Check that column before you write anything; it is the table name you will type.
Step 2 — Look before you leap
Always start by seeing what you actually have:
SELECT * FROM S LIMIT 20
Press Ctrl+Enter to run. Twenty rows appear, and the header row tells you every column name. Column headers containing spaces are converted to underscores, so order date becomes order_date.
Then get the size of the thing:
SELECT COUNT(*) AS row_count FROM S
Step 3 — Ask it questions
This is where SQL earns its keep. Four queries answer most first questions about a dataset.
What distinct values are in a column?
SELECT DISTINCT status FROM S
How many of each?
SELECT status, COUNT(*) AS n
FROM S
GROUP BY status
ORDER BY n DESC
What are the totals, broken down?
SELECT region, COUNT(*) AS orders, SUM(amount) AS revenue
FROM S
GROUP BY region
ORDER BY revenue DESC
Which rows are the extremes?
SELECT * FROM S ORDER BY amount DESC LIMIT 10
Three steps, no setup, and your file never leaves your machine.
Open the tool →Reading the results
Results appear below the editor, paginated at 1,000 rows per page with Previous / Next buttons. If a file was truncated for any reason, a notice appears above the results rather than the rows quietly disappearing.
Click Export to download what you are looking at as CSV, JSON, Excel or Parquet.
Four things that trip people up first time
| Symptom | Cause and fix |
|---|---|
| “Unknown table A” | The alias is not A. It comes from the filename — check the Alias column, or type your own letter there. |
| Text comparison returns nothing | Strings need single quotes: WHERE city = 'Austin', not double quotes. |
| Number comparison returns nothing | Numeric-looking values are detected as numbers, so use WHERE amount > 100 with no quotes. If the column has leading zeros it is kept as text on purpose, so quote it. |
| A parse error mentioning your alias name | Some words are reserved in SQL. AS total will fail; AS total_amount works. |
Where your file goes
Nowhere. Everything above happens inside your browser tab — the CSV is never uploaded. You can prove it in about a minute; see querying without uploading.