Core Concepts

Understanding the foundational concepts of Modality.

Data Modeling Layers

Modality supports three modeling layers:

1. Conceptual Layer

• Purpose: Business view of data
• Components: Domains, Entities, Relationships
• Format: MML domain blocks
• Audience: Business stakeholders, analysts

2. Logical Layer

• Purpose: System implementation view
• Components: Sources, Mappings, Transformations
• Format: MML source and mapping blocks
• Audience: Data engineers, architects

3. Data Products

• Purpose: Consumption layer
• Components: Reports, Metrics, Dashboards
• Format: MML product blocks
• Audience: End users, consumers

Key Terminology

Domain : A logical grouping of related entities (e.g., Sales, Marketing, Platform)

Entity : A business object being modeled (e.g., Customer, Order, Product)

Entity Types : - Dimension: Reference data (Customer, Product) : - Fact: Transactional data (Order, Event) : - Metric: Calculated KPIs (Revenue, Active Users) : - Derivation: Derived tables (Customer 360, Product Catalog)

Relationship : Connection between entities (belongs-to, derived-from, contains)

Source : External system providing data (Database, API, File)

Data Product : Curated data asset with defined consumers and SLAs

Model Lifecycle

1. Design - Create conceptual model
2. Map - Connect to source systems
3. Build - Implement transformations
4. Publish - Create data products
5. Version - Track changes in Git

Best Practices

• Start simple: Begin with core entities
• Use domains: Group related entities logically
• Be explicit: Document relationships clearly
• Version everything: Commit changes regularly
• Review changes: Use pull requests for collaboration

Next Steps

• Conceptual Modeling Guide
• MML Syntax Reference