Ten minutes in — walk through the sample

Step 1: open the sample workspace

A freshly installed TableDI is an empty workspace. The card on the home page says No data in here yet — click Start with sample data.

The home page of an empty workspace, with the Start with sample data button

A few seconds later three things appear in the sidebar:

Under Dashboards: Q3 sales overview
Under Tables: (AA)Sales orders, with (AB)Monthly summary indented under it

That indent is not decoration — it means Monthly summary is calculated from Sales orders. The relationship comes back repeatedly.

Note

That card only appears while the workspace is empty. To install a sample later, use Templates in the sidebar and its Sample data tab; there are three — sales, expenses, stock. Each is ordinary tables and dashboards, deletable at any time.

Step 2: look at the messy detail table

Click (AA)Sales orders in the sidebar. It is one quarter of sales orders, 38 rows, deliberately carrying the flaws a CRM export arrives with: the Region column has stray spacing and mixed capitalisation (North, north, NORTH), Status is written three different ways (paid, Paid, PAID), and two rows have no region at all.

The Sales orders table: raw columns on the left, four fx formula columns on the right

Look at the header row: every column carries a code — AA2, AA3 … AA11. AA is this table's code and the number is the column number. Every formula you write uses them — not "the Region column" but AA4.

The first six are raw data. The four after them carry an fx marker; they are formula columns:

Column	Formula	What it does
Region (clean) `AA8`	`UPPER(TRIM(AA4))`	Strips the spacing, normalises the case
Month `AA9`	`LEFT(AA2, 7)`	Cuts `2026-09` out of `2026-09-28`
Paid? `AA10`	`IF(UPPER(TRIM(AA7)) = "PAID", "Paid", "Unpaid")`	Collapses three spellings into two values
Collected `AA11`	`IF(UPPER(TRIM(AA7)) = "PAID", AA6, 0)`	The amount if it was paid, otherwise 0

Click any formula column's header and the formula bar above shows that column's formula (column codes are displayed as column names there, which reads better).

With Region (clean) selected, the formula bar shows UPPER(TRIM(Region))

Not one character of the raw columns changed. The clean values grew in new columns beside them. That is the basic posture for cleaning data here: keep the original, derive a clean one.

Step 3: look at the Workflow graph

Click Workflow, the leftmost tab above the table.

The Workflow graph: raw columns on the left, formula columns on the right, arrows showing which reads which

Every line is one reference: AA9 Month comes from AA2 Date; AA11 Collected uses both AA7 Status and AA6 Amount. Once you have several tables and a lot of formulas, this graph is the fastest answer to "where does this number come from".

Step 4: switch views

To the right of the Sales orders tab are two more: Product × month and Revenue by region. They are two views of this same table — not two more tables. There is one copy of the data.

The pivot view Product × month: products down, months across, amounts in the cells

To add another way of looking at it, click ＋ View at the right of the tab bar: worksheet, form, bar/line, pivot, stacked, funnel, gauge, metric and more all come from there. Each view keeps its own filter and sort, and none of it touches the table.

Step 5: look at a summary table that grew

Click (AB)Monthly summary in the sidebar. Five columns, three rows, and not one number was typed:

Column	Formula	Meaning
Month `AB1`	`AA9`	References the Month column of Sales orders, which yields one row per month
Booked `AB2`	`SUMIF(AA6, AB1 = AA9)`	Sum of amounts in Sales orders whose month equals this row's month
Collected `AB3`	`SUMIF(AA11, AB1 = AA9)`	Same, but summing the collected amount
Orders `AB4`	`COUNTIF(AA2, AB1 = AA9)`	How many orders this month
Average order `AB5`	`AB2 / AB4`	Two columns of this same table, divided

The Monthly summary table: every column is a cross-table formula

This is TableDI's version of a pivot table, with one difference: it is a real table. You can keep adding columns to it, other tables can reference it, and it can be charted. Paste next month's orders into Sales orders and a 2026-10 row appears here on its own.

Step 6: open the dashboard

Click Q3 sales overview in the sidebar.

The dashboard: a heading, a booked-vs-collected bar chart, a pie, a pivot and a detail table

A dashboard stores no data of its own — it places views from several tables side by side and adds text and controls. Edit in the top right enters edit mode, where each block can be dragged and resized and new charts or text added from the right; changes save themselves.

Step 7: change one number and watch them all move

Go back to (AA)Sales orders, double-click any row's Amount and make it obviously different — turn 9,600 into 99,600. Then look:

Booked and Average order for that month changed in Monthly summary;
the bar chart and pie changed on the dashboard;
the Product × month pivot changed.

You did not drag a formula, refresh a pivot table or rebuild a chart. That minute is the whole difference between this and Excel: there is one copy of the data, and everything after it is calculated from that copy.

Wrapping up

The sample deletes like anything else: hover a row under Recently on the home page, click the ⋮ in the Actions column, choose Delete — or right-click it in the sidebar. Deleting the sample dashboard gives the free plan's one dashboard slot back.
The next page shows you how to bring in your own Excel. Bring your messiest file; that is what this is for.