File Upload

Upload CSV or XLSX files to import data into your data warehouse. The platform analyzes your file, detects column types, identifies sensitive data, and lets you review everything before ingesting.

How to Upload

1. Navigate to Data in the sidebar
2. Click the Upload tab
3. Drag and drop your file or click to browse

Supported Formats

Format	Extensions	Notes
CSV	.csv	Auto-detects delimiter (comma, semicolon, tab, pipe)
Excel	.xlsx	Multi-sheet support — you choose which sheet to import

File Size Limits

File size limits depend on your plan:

Plan	Max File Size
Free	10 MB
Starter	100 MB
Growth	500 MB
Business	2 GB

Analysis Phase

After you select a file, the platform automatically analyzes it:

What Gets Detected

• Column names and suggested clean names
• Data types: string, number, date, boolean, datetime
• PII flags: Email addresses, phone numbers, CPF/CNPJ, credit card numbers
• Sheet selection (for XLSX with multiple sheets)

Dataset Matching

If the platform finds an existing dataset with >60% column name match, it will ask:

"We found a dataset that looks similar: sales_data. Would you like to merge into it or create a new dataset?"

This prevents accidental duplicates and makes it easy to append new data to existing tables.

Review & Configure

Before ingesting, you can review and customize:

Column Configuration

For each column, you can customize:

Setting	Description
Name	Rename the column for clarity
Type	Override the detected type (string, number, date, boolean, datetime)
Description	Add a human-readable description
PII	Mark as personally identifiable information (will be masked in queries)
Key	Mark as a primary/business key column
Exclude	Skip this column during ingestion

Available Transformations

You can apply transformations to columns before ingestion:

Transformation	What It Does	Example
Extract First Word	Takes the first word from a string	"John Doe" → "John"
Lowercase	Converts to lowercase	"HELLO" → "hello"
Uppercase	Converts to uppercase	"hello" → "HELLO"
Trim	Removes leading/trailing whitespace	" hello " → "hello"
Monetary (BRL)	Parses Brazilian currency format	"R$ 1.234,56" → 1234.56
Excel Date → Datetime	Converts Excel serial dates	45292 → "2024-01-15"
Boolean (PT-BR)	Converts Portuguese boolean strings	"Sim"/"Não" → true/false

Ingestion Mode

Choose how data is written:

Mode	Behavior
Append	Add new rows to the existing dataset (default)
Replace	Delete all existing rows and write the new data
Merge	Upsert: update matching rows, insert new ones (requires key columns)

Dataset Description

Add a description for your dataset. This helps the AI understand your data better when answering questions:

"Monthly sales transactions from our e-commerce store. Each row is a sale with product, quantity, price, customer region, and payment method."

Ingestion Result

After clicking Ingest, you'll see a success report:

Field	Example
File	sales_jan_2025.csv
Target Table	sales_data
Rows Ingested	15,234
Schema	tenant_abc123
Mode	Append

Your data is now available for AI Chat, Dashboards, KPIs, and Analysis.

Tips for Best Results

info

Best Practices for Data Upload

1. Use descriptive column headers — "revenue_usd" is better than "col1"
2. Include dates — time-series data enables trend analysis and forecasting
3. Add descriptions — help the AI understand what each column represents
4. Mark PII columns — the platform will auto-mask sensitive data
5. Start with clean data — remove blank rows and fix formatting before upload