Skip to main content

Uploading Files

In addition to connecting live data sources, Datafi lets you upload flat files directly into the platform. This is useful for one-off analyses, reference tables, lookup data, or any dataset that exists as a local file rather than a live database.

Supported File Formats​

Datafi supports two file formats for upload:

  • CSV -- Comma-separated values. The most common format for tabular data exports.
  • JSON -- JavaScript Object Notation. Suitable for semi-structured or nested data.
Other Formats

If your data is in Excel, Parquet, or another format, convert it to CSV or JSON before uploading. Most spreadsheet applications and data tools support CSV export.

How to Upload a File​

  1. Open the Data Catalog and click Add Dataset.
  2. Select the CSV or JSON connector tile.
  3. Drag and drop your file into the upload area, or click the area to browse your local filesystem and select a file.
  4. Datafi reads the file and displays a schema preview showing detected columns, data types, and sample rows.
  5. Review the schema. Adjust column names or data types if the auto-detection is not accurate.
  6. Provide a Dataset Name that will identify this file in the catalog.
  7. Click Upload to finalize.

The uploaded file is stored within Datafi and becomes a queryable dataset, just like a connected database. You can query it, join it with other sources, share it, and apply policies to it.

Drag and Drop

The upload area supports drag and drop from your operating system's file manager. Simply drag the file from your desktop or file explorer and release it over the upload zone -- no need to click through a file picker.

File Format Requirements​

CSV​

  • The first row must contain column headers.
  • Use commas as the delimiter. Tab-separated or semicolon-separated files are not automatically detected.
  • Enclose values that contain commas, newlines, or double quotes in double quotes.
  • Use UTF-8 encoding.

JSON​

  • The file must contain a valid JSON array of objects, where each object represents a row.
  • All objects should share a consistent set of keys. Missing keys are treated as null values.
  • Nested objects are flattened during import. Deeply nested structures may not map cleanly to tabular columns.
  • Use UTF-8 encoding.

Storage Limits and Pricing​

Datafi provides a free storage tier for uploaded files. You can upload and store files at no additional cost up to the limit included with your plan.

If your total uploaded file storage exceeds the free tier:

  • Additional storage is billed at the rate specified in your plan.
  • You receive a notification when you approach your storage limit.
  • Existing files remain accessible even after you exceed the limit, but new uploads may be blocked until you free up space or upgrade.
Connected Sources Are Not Affected

Storage limits and pricing apply only to uploaded files. Connected data sources -- databases, warehouses, and SaaS platforms -- have no size limit and are not charged based on data volume. You pay for query usage, not for the amount of data in the source.

Managing Uploaded Files​

Once uploaded, a file behaves like any other dataset in the Data Catalog. You can:

  • Browse the schema -- View columns, data types, and sample data from the dataset details page.
  • Query the data -- Use the Datafi query editor or natural-language chat to run queries against the uploaded file.
  • Share access -- Invite internal users or external partners, just as you would with a connected source.
  • Apply policies -- Define row-level and column-level access rules.
  • Delete the file -- Remove the uploaded dataset permanently. See Managing Datasets for deletion instructions.
Re-Uploading Data

If the underlying data changes, you need to upload the updated file again. Uploaded files are static snapshots -- they do not sync with a source automatically. For data that changes frequently, consider connecting the source directly instead of uploading files.