Skip to content
M↓ View as Markdown ↗

CSV files

Uploading a CSV file is a quick way to try Actian AI Analyst with your own data. It suits demos, prototypes, and tests. It isn't meant as a long-term or production data source: an uploaded file doesn't change unless you upload it again. For live analysis, connect your data warehouse instead (see Connect a data source).

Availability

  • Who can do this: members with the Edit capability in the active Workspace. If Restrict connection management is on, you also need the Manage organization capability.
  • Deployment: Actian SaaS and self-managed deployments.

Prepare your file

  • Save the file as .csv, with column headers in the first row.
  • Use UTF-8 encoding.
  • The delimiter, header row, and column types are detected automatically. Commas are the safest delimiter.
  • Don't upload Excel workbooks (.xlsx or .xls). Export the sheet to CSV first. Studio's file browser only offers .csv files. A different file that you drag in is still uploaded, but processing usually fails.
  • Avoid formulas, merged cells, multiple header rows, and blank rows or columns.
  • Use clear, consistent column names and consistent date and number formats.

Upload a file

  1. In Studio, open Connections and click + (Add connection).
  2. Under Files, choose CSV file. The Upload a CSV page opens.
  3. Drop a CSV file onto the upload area ("Drop a CSV here or click to browse"), or click the area and pick a file. The upload starts right away and there is no separate confirm button. Studio then opens the file's page.
  4. Wait for the status to change from Processing to Ready. The page updates by itself. During processing, column types are inferred.

Upload one file at a time. If you drop several files, only the first one is uploaded. The upload area shows the size limit ("CSV files up to N MiB"). The default is 250 MiB, but your administrator may have set a different limit. A file over the limit is rejected with "This file is too large to upload".

Check the result

The file's page has two tabs:

  • Overview: the File summary shows the file's status (Processing, Ready, or Failed), Rows, and Columns.
  • Columns: each Column with its inferred Type. Types use the database's own names, such as VARCHAR, BIGINT, DOUBLE, or DATE.

Uploaded files are not included in the daily connection health checks.

How names are derived

AI Analyst turns the file name into a table name, and each header into a column name:

  • The file extension is removed (only the last one, so a.b.csv becomes a_b).
  • Letters are lowercased.
  • Every run of characters other than a–z and 0–9 becomes _. This includes spaces, punctuation, and accented letters.
  • Leading and trailing underscores are removed.

For example, Sales Q1.csv becomes the table sales_q1. If nothing is left, the table is named file and the column is named column. If two columns end up with the same name, the second gets _2, the third _3, and so on.

File names must be unique within a Workspace after this conversion, so Sales Q1.csv and sales-q1.csv count as the same file. A second upload with the same name is rejected, even if the earlier file failed. To replace a file, delete the old one and upload the new one.

Use the file in an AI Analyst

AI Analysts answer questions through Models, not raw files. When a file is Ready:

  1. Create a Model on it. When you create the Model, choose CSV Uploads under Data connection, then pick the file's table. See Creating models from a table.
  2. Add that Model to an AI Analyst.

If your data is relational, upload each table as its own file, for example:

  • customers.csv with customer details
  • orders.csv with order details
  • products.csv with the product catalog

All ready files in a Workspace can be queried together. AI Analyst doesn't infer how the files relate, so define the joins between their Models as Relationships.

Where uploaded files live

Each uploaded file belongs to the Workspace you uploaded it in, and Studio lists it under Files in Connections. It can't be shared with other Workspaces. To use the same data in another Workspace, upload the file there too. See Data connections and Workspaces.

Delete a file

  1. Open the file's page from Connections → Files.
  2. Click the trash icon (Delete file).
  3. In Delete file?, click Delete.

Deleting permanently removes the uploaded file and its stored data. This cannot be undone. Before you delete a file, update or delete any Models built on it.

Troubleshooting

Message Where What to do
This file is too large to upload Upload area Split the file, or ask your administrator about the limit.
Could not upload file Upload area Check that no file with the same name already exists in this Workspace, and that you have permission to upload. Then try again.
The uploaded file is empty File page The file has no content. Check the export.
The uploaded file could not be parsed as CSV File page The file isn't valid CSV, for example an Excel workbook. Export it as CSV and upload it again.
The uploaded file has no data rows File page The file has a header but no rows.
File processing failed File page Any other processing error. Check the file's formatting and encoding, then delete it and upload it again.

If the problem continues, contact your administrator or contact support. Where available, you can also use Support in Studio's sidebar.