Introduction
Every business drowns in spreadsheets—endless rows of customer information scattered across files, employee records buried in different systems, project data trapped in email chains and shared drives. Teams waste hours hunting for the right version, copying data between systems, and fixing errors that multiply when information gets manually transferred. Meanwhile, critical insights hide in plain sight because your data lives in isolation, disconnected from the workflows that could transform numbers into action.
Proma Datasets transform this chaos into intelligent, interconnected information systems. Unlike static spreadsheets that just hold values, Datasets are living containers where data validates itself, relationships form automatically, and business logic embeds directly into your information structure. It's about creating a foundation where your data works as hard as your team does—preventing errors before they happen, connecting information across your organization, and turning raw facts into actionable intelligence.
What is a Proma Dataset?
A Proma Dataset is an intelligent table that stores your structured business information with embedded logic, validation, and relationships. Think of it as a spreadsheet that evolved—where columns understand their content, enforce business rules, and connect seamlessly with other datasets across your organization.
Datasets serve as the foundation for every system you build in Proma. Whether tracking employees, managing projects, or monitoring inventory, each dataset becomes a specialized repository that not only stores information but actively maintains data quality and enables sophisticated business logic.
Getting Started
Accessing Datasets
Datasets live within systems in your Proma workspace. Navigate to any system and toggle Build Mode to create and manage datasets. Each system can contain multiple datasets designed for specific purposes—customers, orders, tasks, or any business entity that requires structured tracking.
When no datasets exist in a system, you'll see a "No datasets found" message with a + New Dataset button. The dataset interface presents your information in familiar table format while providing powerful tools for filtering, sorting, and bulk operations.
Creating Your First Dataset
Dataset creation begins with AI-powered assistance that understands your business needs:
Step 1: Name and Describe Start by giving your dataset a meaningful name and description. The AI analyzes your description to suggest relevant column structures—write something like "create a to do list app" and watch the intelligence unfold.
Step 2: AI Column Suggestions Based on your description, Proma's AI suggests appropriate columns with smart defaults:
Task name (Text column)
Priority (Options column with predefined choices)
Start Date and Due Date (Date columns)
Notes (Multi-line text column)
Status (Options column for workflow tracking)
Step 3: Customize and Add Review AI suggestions, modify column types as needed, and add additional columns using the + Add Column interface. Each column type brings specialized functionality—from simple text storage to complex approval workflows.
The AI learns from your choices, providing increasingly relevant suggestions as you build more datasets within your organization.
Understanding Dataset Structure
Datasets consist of rows (individual records) and columns (data fields), but with significantly more intelligence than traditional tables:
Rows represent individual instances of whatever you're tracking—specific employees, particular projects, or unique customer orders. Each row maintains relationships with other datasets and triggers automations based on changes.
Columns define not just what data to store, but how it behaves, validates, and interacts with other information. Smart columns embed business logic directly into your data structure, reducing complexity in views and automations.
Core Components
Datasets integrate seamlessly with Proma's core building blocks to create intelligent, automated business systems:
Smart Columns - The foundation of every dataset, these intelligent fields go far beyond simple data storage. Smart columns embed validation, calculations, and specialized functionality directly into your data structure. From basic text and numbers to advanced approval workflows and payment processing, each column type brings purpose-built intelligence to your information.
Automation Engine - Datasets come alive through automated workflows that respond to data changes. When records are created, updated, or deleted, automations can trigger notifications, update related information, or execute complex business processes. Every dataset interaction can spark intelligent responses across your entire system.
Logic Builder - Embedded business rules and calculations operate directly within your dataset structure. Set validation logic to enforce data quality, create calculated fields that update automatically, and implement conditional formatting that provides visual insights. Logic Builder ensures your data behaves according to your business requirements without manual intervention.
These components work together to create datasets that don't just store information—they actively participate in your business processes, maintaining quality, triggering actions, and providing intelligence exactly where you need it.
Common Mistakes to Avoid
Dataset Design Mistakes
Over-Engineering from the Start
New users often try to build comprehensive datasets that handle every possible scenario immediately. This leads to overly complex structures that are difficult to use and maintain. Start with core fields that handle 80% of your use cases, then expand gradually as real needs emerge.
Duplicating Data Across Datasets
Storing the same information in multiple datasets creates synchronization nightmares and data inconsistencies. Use lookup columns to reference shared information from a single source of truth rather than copying data between datasets.
Ignoring Column Type Selection
Choosing generic text columns for everything sacrifices validation, automation capabilities, and user experience. A phone number stored as text can't provide click-to-call functionality, and dates stored as text can't be filtered or sorted properly.
Naming and Organization Mistakes
Inconsistent Naming Conventions
Using different naming patterns across datasets creates confusion and makes automation development difficult. Establish clear conventions for dataset names, column names, and option values before you start building.
Generic or Unclear Names
Datasets named "Data," "Items," or "List" become impossible to distinguish as your system grows. Use specific, descriptive names that clearly indicate the dataset's purpose—"Customer Contacts," "Project Tasks," or "Inventory Items."
Validation and Logic Mistakes
Skipping Validation Rules
Datasets without proper validation allow bad data entry that creates problems throughout your system. Required fields, format validation, and business rule enforcement prevent data quality issues before they start.
Integration and Automation Mistakes
Building Datasets in Isolation
Datasets that don't connect with your broader system miss opportunities for automation and workflow integration. Consider how each dataset fits into your larger business processes from the beginning.
Overcomplicating Relationships
Complex many-to-many relationships and deeply nested lookups can create confusing user experiences and performance problems. Keep relationships simple and intuitive—users should understand how data connects without extensive training.
Building Effective Datasets
Design Principles
Start with Business Entities - Identify the core objects your organization tracks—customers, projects, employees, products. Each significant business entity typically warrants its own dataset with relationships connecting related information.
Embed Logic at the Data Level - Use column intelligence to enforce business rules consistently. Validation logic prevents bad data entry while calculation fields automate routine computations without user intervention.
Design for Growth - Structure datasets to accommodate expanding information needs. Generic text fields can evolve into specialized columns while maintaining data integrity during structural changes.
Data Organization Strategies
Normalize Relationships - Avoid duplicating information across datasets. Use lookup columns to connect related data while maintaining single sources of truth for each piece of information.
Balance Detail and Usability - Include sufficient detail for operational needs without overwhelming users with unnecessary complexity. Hidden or optional fields can accommodate edge cases without cluttering standard interfaces.
Plan for Integration - Design datasets with external system connections in mind. Consistent naming conventions and data formats simplify automation development and third-party integrations.
Adoption and Training
Gradual Implementation - Introduce datasets incrementally rather than replacing entire information systems simultaneously. Teams can adapt to new capabilities while maintaining operational continuity.
Role-Based Access - Configure initial permissions that match existing organizational responsibilities. Users can immediately work with relevant information while maintaining appropriate security boundaries.
Troubleshooting
Common Data Issues
Validation Failures
Users sometimes encounter validation errors when entering information that doesn't conform to business rules. Review column configuration to ensure validation logic matches actual business requirements rather than theoretical ideals.
Performance with Large Datasets
Datasets with thousands of records can experience slower loading times. Implement appropriate filtering defaults and consider archiving historical data that doesn't require immediate access.
Import Problems
Data imports from external systems may fail due to format mismatches or validation conflicts. Prepare import files using dataset structure guidelines and validate sample data before processing large batches.
Relationship and Integration Issues
Broken Lookup Connections
Lookup columns may lose connections if referenced records are deleted or moved. Implement deletion policies that protect against relationship breaks or provide alternative reference options.
Permission Conflicts
Complex permission configurations can create situations where users cannot access expected information. Review the complete permission hierarchy from organization through spaces to specific dataset columns.
Synchronization Problems
External system integrations may create data inconsistencies during network outages or service disruptions. Monitor integration health and implement retry logic for critical data synchronization processes.
Next Steps
Explore related Proma features that enhance dataset capabilities:
@Automation Engine - Create workflows that respond to dataset changes and maintain data consistency across your organization
@Logic Builder - Implement complex business rules and calculations that operate directly within your data structure