Home› Getting Started› Quick Start Guide› Run Data Warehouse Wizard

Run Data Warehouse Wizard

After connectors and metadata are available in the repository, the next step is to generate a draft data warehouse model using the AnalyticsCreator wizard. This is the central step where the system translates metadata into a structured warehouse design.

The wizard analyzes imported metadata and automatically creates a full model including staging, historization, and transformation layers. This provides a working baseline that can be refined instead of built manually from scratch.

Purpose

Generate a complete draft data warehouse model based on imported metadata, including STG, CORE, and DM structures.

Design Principle

AnalyticsCreator follows a generation-first approach:

  • The full data warehouse model is generated automatically from metadata
  • Developers refine and adjust the generated model instead of building it manually

The wizard uses structural metadata such as tables, keys, and relationships to infer joins, dimensions, and fact structures.

Inputs / Outputs

Inputs

  • Imported metadata from connectors
  • Selected source tables
  • Modeling approach (e.g. Data Vault, dimensional, mixed)
  • Optional configuration (naming conventions, defaults)

Outputs

  • Generated data warehouse model including:
    • STG layer (import structures)
    • Persistent staging and historization structures
    • CORE layer (dimensions and facts)
    • DM layer (analytical structures)
  • Predefined joins and relationships
  • Initial transformation logic

Internal Mechanics

1. Metadata analysis

The wizard reads all metadata stored in the repository, including tables, columns, and relationships. Based on this, it determines how objects are related.

2. Model generation

AnalyticsCreator generates a complete data warehouse structure. This includes:

  • Import tables in the STG layer
  • Persistent staging structures with historization
  • CORE transformations for dimensions and facts
  • DM structures for analytical consumption

3. Relationship inference

Joins between tables are derived automatically based on source relationships. These joins are used to construct fact and dimension transformations.

4. Default logic generation

The wizard can apply default behaviors such as:

  • Including directly and indirectly related tables in facts
  • Creating standard transformations
  • Generating calendar dimensions

5. Visual model creation

The result is a fully structured data warehouse diagram that shows all layers and dependencies. At this stage, the model is defined but not yet deployed.

Types / Variants

Modeling approaches

  • Data Vault model (hubs, links, satellites)
  • Dimensional model (facts and dimensions)
  • Mixed approach (Data Vault foundation with dimensional output)

Configuration options

  • Naming conventions (prefixes, suffixes)
  • Default transformations
  • Inclusion rules for related tables

Example

A set of source tables is selected:

  • Customer
  • Orders
  • OrderLines
  • Products

After running the wizard:

  • STG tables are created for each source
  • Relationships are detected automatically
  • A fact table is generated based on transaction data
  • Dimensions are generated for related entities

The resulting model already contains joins, transformation paths, and structural dependencies.

When to Use / When NOT to Use

Use when

  • Starting a new data warehouse model
  • Rapidly generating a baseline structure
  • Working with well-defined source metadata

Do NOT rely on defaults when

  • Business logic is complex or non-standard
  • Source relationships are incomplete or incorrect
  • Fact and dimension definitions require domain-specific adjustments

Performance & Design Considerations

  • The wizard accelerates initial modeling but does not replace design decisions
  • Generated joins should be reviewed for correctness and performance
  • Fact table scope depends on inclusion settings (direct vs indirect relationships)

Design trade-off:

  • Full automation provides speed
  • Manual refinement ensures correctness and performance

Integration with other AnalyticsCreator features

  • Repository: provides metadata input for the wizard
  • Transformations: generated and refined after wizard execution
  • Synchronization: converts generated model into SQL objects
  • Deployment: packages generated artifacts

Common Pitfalls

  • Assuming the generated model is production-ready without review
  • Over-including tables leading to overly complex fact structures
  • Ignoring incorrect or missing source relationships
  • Not validating generated joins

Key Takeaway

The wizard generates a complete data warehouse model from metadata, which is then refined and deployed rather than built manually.