Input Data
Drag Input Data, then browse to a file or set up a connection. It reads CSV, Excel, JSON, XML, and most databases via ODBC or a native connector.
For Excel, you choose the sheet or a named range. For CSV, the settings that matter are:
- Field delimiter — comma, tab, pipe
- First row contains field names — usually yes
- Start data import on line — skips title rows
- Code page — the fix when accented characters come through wrong
Reading many files at once
In the file path, use a wildcard: C:\data\sales_*.csv.
Set Output File Name as Field to add a column recording which file each row came from — essential when you are combining twelve monthly extracts and need to know which is which.
All files must share the same structure. If they do not, read them separately and union with the Union tool, which can match on field name.
Output Data
Writes to file or database. The setting to get right is what happens to an existing target:
| Option | Effect |
|---|---|
| Overwrite file | Replaces it entirely |
| Append to existing | Adds rows — duplicates if the workflow reruns |
| Overwrite table (drop) | Drops and recreates, losing indexes and permissions |
| Delete data and append | Keeps the table structure, replaces the rows |
Append to existing is the usual cause of duplicated data: run the workflow twice and you get every row twice. If the workflow may be rerun, use delete-and-append or overwrite.
The Browse tool
Browse shows the full data at a point in the workflow, with column profiles — data types, null counts, value distributions.
Useful during development, but each one holds the whole dataset in memory. Remove or disable them before running against production volumes.