Skip to content

File Uploads (CSV, Excel, JSON)

AnalytAI allows you to turn static files like CSV, Excel, and JSON into a powerful, interactive database. Instead of manually filtering rows in a spreadsheet, you can simply ask questions and get instant answers.


You can mix and match multiple files in a single project:

  • CSV (.csv)
  • Excel (.xlsx, .xls)
  • JSON (.json)

When you upload files, AnalytAI automatically:

  1. Analyzes the structure of your files.
  2. Creates a database table for each file.
  3. Detects column types (Dates, Numbers, Text).
  4. Connects the data so you can ask cross-file questions (e.g., “Compare sales in this CSV to targets in this Excel”).

  1. Go to the Add Data Source section in AnalytAI.
  2. Select File-Based Sources.
  3. Choose your Organization and give your project a name (e.g., “Marketing Analytics 2024”).
  • Drag and drop your files or click to browse.
  • You can upload multiple files at once.
  • Tip: Make sure your files have clear header rows (column names).
  • AnalytAI will show you a preview of each file.
  • You can rename the Table Name (this is how you will refer to the data in chat).
  • If a file has errors, AnalytAI will highlight them so you can fix your spreadsheet and re-upload.
  • Click Upload to Database.
  • Once the progress bar completes, click Setup Project.
  • You’re ready! Start asking questions like “What was the average order value from the sales.csv file?”

  • Clean Headers: Ensure the first row of your CSV/Excel contains clear names (e.g., Order_Date instead of Col1).
  • Consistent Dates: Use standard date formats (YYYY-MM-DD or DD/MM/YYYY).
  • JSON Structure: For JSON files, ensure data is structured as an array of objects for best results.

  • File Too Large? For very large datasets (over 50MB), we recommend using a SQL database connection instead.
  • Unsupported Format? Ensure your Excel file isn’t password-protected.
  • Wrong Data Types? If a number column is being treated as text, check if there are any non-numeric characters (like “N/A” or “TBD”) in that column.

Return to Connectivity