Skip to main content

​
Creating a Dataset

Click Datasets in the sidebar → Create Dataset. Dataset Name: Choose descriptive names like “Customer Contact Data” or “Product Inventory Q4”. Avoid generic names like “Dataset 1”. Description: Explain what data the dataset contains, its source, and intended use. 
Customer contact information collected from website sign-ups
January-March 2024. For lead qualification and sales outreach.

​
Column Planning

Plan your column structure before creating the dataset. Consider:
  • Unique Identifier: Column that uniquely identifies each record
  • Primary Data: Main information to process
  • Context Columns: Additional supporting information
  • Output Columns: Where processed results are stored
You can start empty and add columns later, or import data with columns defined from CSV headers.

​
Adding Columns

Click Add Column to configure each field: Column Name & Key: Display name (e.g., “Company Name”) and system identifier (e.g., “company_name”) Data Type: Short Text, Long Text, Email, URL, Phone, Number, Date, Boolean, or JSON Description: What the column contains and how it should be used AI Instructions: (Optional) How agents should process this column Validation Rules: Data quality requirements

​
Example Column

Name: Company Name
Key: company_name
Type: Short Text
Description: Official company name from website or registration
Instructions: Extract official name, remove LLC/Inc unless part of brand
Required: Yes

​
Importing Data

After creating your dataset structure, import data via CSV/Excel upload. The system will map file columns to your dataset columns and validate data during import.
You can also create records manually for small datasets or testing.

​
Next Steps

Import Data

Import CSV files and other data sources

Manage Columns

Configure column types and validation

Create AI Agents

Build agents to process your dataset

Work with Records

Add and manage individual records