Skip to main content

​
Column Configuration

Open your dataset → Click column header → Edit Column. Column Name & Description: Use descriptive names and explain the column’s purpose. Data Type: Short Text, Long Text, Number, Date, Boolean, Email, URL, Phone, or JSON. Validation Rules: Set length limits, format patterns, value lists, min/max ranges, or conditional rules based on other columns.

​
Example Validation

Phone Numbers: ^\+?1?\d{9,15}$
Industry Values: Technology, Healthcare, Finance, Manufacturing, Retail, Other
Conditional: If Country = "US", then State must be valid US state code

​
AI Agent Assignment

Assign agents to columns for automatic processing: Enrichment Mode: Agent adds information to empty cells, preserving existing data. Enhancement Mode: Agent improves or standardizes existing data. Analysis Mode: Agent analyzes data and adds insights, original data preserved.

​
Column Instructions

Customize agent behavior with column-specific guidance: 
For "Company Name":
- Extract official business name
- Remove LLC/Inc unless part of brand
- Standardize capitalization
- Handle abbreviations consistently

​
Processing Settings

Batch Size: Records to process at once (default: 100) Schedule: Immediate, scheduled, triggered, or manual processing

​
Data Quality

The system calculates quality metrics for each column: Completeness: % of non-empty values Validity: % passing validation rules Consistency: Standardization and format consistency Accuracy: Estimated accuracy from validation Set thresholds to trigger alerts when quality drops below acceptable levels.

​
Next Steps

Working with Records

Edit, filter, and manage data records

Creating Agents

Build AI agents to process column data

Importing Data

Import data into your datasets

API Reference

Technical documentation for columns