Norami DocsAdding your data
Upload spreadsheets, confirm how they're read, and import them into Norami.
Your assistant answers from data you give it. The first step in building an assistant is uploading a spreadsheet and confirming Norami read it correctly.
Supported files
Norami imports tabular files:
- Excel —
.xlsx,.xls - CSV —
.csv
Each file should have a header row (the row with your column names) and rows of data beneath it.
The upload flow
Choose Add data
From your workspace, select Add data and pick your file.
Norami detects the structure
Norami scans the first rows to find the header row and identify each column. For workbooks with multiple sheets, it handles each sheet it can read.
Confirm the preview
You'll see a preview of how your data was understood — the detected headers and sample rows. Check that columns line up with the right headers.
Import
Confirm to import. Norami processes the file and makes the data available to your assistant.
Tips for clean imports
- Keep one header row. Multiple stacked header rows or merged cells can confuse detection. A single, clear header row imports most reliably.
- Use real column names. "Revenue" reads better than "Column F."
- Consistent types per column. A column that mixes numbers and text (e.g. amounts mixed with notes) is harder to analyze.
- Remove decorative rows like titles or totals above the header, when you can — though Norami tries to skip them automatically.
If the preview looks wrong — the header was detected on the wrong row, or a column is mislabeled — adjust it on the preview screen before confirming. That's the moment to fix mapping issues.
What happens next
After import, Norami sets up an assistant from your data automatically. Continue to How Norami sets up your assistant.
Adding more later
You can add more files to an existing assistant, or refresh a dataset with a newer version of the same file. See Updating & refreshing data.