Prepare a table
Read CSV, Excel, JSON and other supported files. Choose a source format and, optionally, a worksheet, delimiter or header row.
Browse all 27 workflow actions. Prepare a file, pass its table content through your actions, and export the result.
SQL and workbook preparation require paid access on your AltDat plan. Microsoft connector licensing is separate. Read about workbook credits.
27 actions
Bring files in, return rows to your flow and save your results.
Read CSV, Excel, JSON and other supported files. Choose a source format and, optionally, a worksheet, delimiter or header row.
Read multiple worksheets as separate tables or combine them into one table with worksheet names.
Return a table to your flow as one record per row, for Apply to each, creating items or sending emails.
Save an AltDat table as CSV, Excel, JSON, Parquet or another supported format.
Choose, organize and summarize your data.
Keep rows that match a condition on a column.
Put rows in ascending or descending order.
Keep the columns you need in the order you choose.
Give a column a new name while preserving its values.
Keep one matching row or only rows with no duplicates, using selected columns or the full row.
Replace missing values with a chosen value or supported fill method.
Convert a column to the data type needed by the next action.
Group matching values and count rows or summarize a value column.
Turn values from one column into columns for a summary table.
Apply an ordered set of operations in one action using a JSON pipeline.
Extract values, compare text and expand lists.
Extract the first regular-expression match or capture group into a new column.
Trim whitespace or change letter case, keeping the original column.
Replace matching text and write the result into a new column.
Count the edits needed to make two values match. A distance of zero means equal.
Score two text values with Jaro-Winkler similarity. One means equal; lower scores mean less similar.
Read a JSON path from a column into a new text column. Missing values return null.
Make a row for each list item and repeat the other columns alongside it.
Match records or stack datasets.
Match rows from two prepared tables using a shared key column.
Stack a second prepared table underneath the first.
Inspect a table and answer questions.
Get ordered column names, row and column counts, and numeric summary statistics.
Compare relationships between numeric columns with a correlation matrix.
Inspect per-column distributions and summary statistics.
Run one read-only SQL query against the prepared table named data.