Rowmotive help
From your first row to your export.
Bring a table
Copy cells from a spreadsheet, including a header row, then paste them into Rowmotive. You can also upload or drop a CSV, TSV, or semicolon-separated file up to 5 MB. Rowmotive imports up to 10,000 rows and 100 columns. It repairs common problems instead of rejecting the file: repeated or blank column names, rows with missing or extra values, stray quotes, and blank lines. A note under the toolbar lists every repair.
Start from suggested AI columns
After you import a table, Rowmotive describes each column and suggests up to five AI columns, such as a category, a yes-or-no check, or an urgency score. It fills the first 5 rows of each suggestion right away so you can judge them. Keep the useful ones and remove the rest. Suggestions read your headers and a sample of up to 30 rows.
Add your own AI columns
Use the plus button in any column header to add an AI column right after it, or Add AI column to put one at the end. You can have up to 10. Click an AI column header to change its name, answer type, instructions, position, and the columns it reads. Changing its task clears its answers.
Choose what each AI column reads
A new AI column reads every column, each labeled with its header. Untick IDs, emails, or anything the answer does not need; unticked columns are not sent for that AI column but stay in your download. Rows where every ticked column is blank are skipped. Each row may send up to 4,000 characters.
Write a specific task
For Category, name 2-20 distinct categories. Add Other if some rows may not fit. For Yes or no, ask a single question that the supplied text can answer. For Score, write each level on a separate line, from lowest to highest. Scores begin at 1.
Check the first answers
Each new or changed AI column fills its first 5 rows automatically. Check them before filling the rest. If you change a task, its answers are cleared so answers from different instructions do not get mixed.
Fill and review
The fill bar shows how many answers the remaining rows need before you click Fill AI columns. You can stop after the current batch. Some answers are flagged for review, but unflagged answers can still be wrong. Check the results against your source text and click a result field to correct it. For an explicit missing-information answer, use Category with Yes, No, and Not enough information, and define those categories in your instructions.
Export before leaving
Download the complete table with each AI column in place, plus a review column for any AI column that flagged a row. You can also copy all AI columns, with their headers, back into your spreadsheet. Original order is preserved. Source data lives in the current tab and is cleared when you leave or refresh. CSV export prefixes potentially executable spreadsheet formulas with an apostrophe.
Understand your allowance
An answer is one AI column answering one nonblank row, so three AI columns on 50 rows use 150 answers. The free allowance is 250 lifetime answers once live processing is available. The proposed Pro offer includes 50,000 per UTC calendar month for $29 per month. The automatically answered first rows count. No automatic overages are charged. Checkout is available only when billing is configured.
Use suitable data
The AI evaluates your supplied text. It does not visit links or verify facts. Do not submit passwords, payment card details, health records, or data you are not authorized to share with the processing providers. Review answers before acting on them.
Sensitive topics and blocked uses
AI columns about health, mental health, money and tax, legal matters, or decisions about people are filled normally and show a disclaimer: their answers are general information, not professional advice. Before an AI column is filled, an automated check reads its task and blocks a narrow set of harmful uses, such as planning violence or stalking someone. A blocked column uses no answers. The content policy lists what is blocked and what stays allowed.
Need a hand? Contact support. Share a small fictional example rather than private customer data.
Try a sample table