Timeslice-Based History: Preserving Data Accuracy Over Time
How the Nexus Model Architecture handles temporal data changes with automatic historical tracking, enabling accurate point-in-time analytics and compliance reporting.
The Challenge of Time in Data Warehousing
Business data changes over time. The Nexus Model Architecture treats time as a first-class dimension, enabling accurate historical reporting and trend analysis without the complexity of traditional slowly changing dimensions.
In this article, we'll explore how the Nexus Model Architecture handles temporal data changes through its timeslice mechanism, providing automatic historical tracking for point-in-time analytics and compliance reporting.
Changes Over Time"] --> B["Customer Status
Active to Inactive"] A --> C["Product Price
99 to 119"] A --> D["Contract Terms
3y to 5y"] B --> E["Timeslice 1
2023-01-01 to
2024-03-15"] C --> F["Timeslice 2
2024-03-16 to
9999-12-31"] E --> G["Point-in-Time Query
Status on 2023-06-01"] F --> G G --> H["Accurate Historical
Analytics & Reporting"] style A fill:#e3f2fd,stroke:#1976d2 style H fill:#e8f5e9,stroke:#388e3c
The Timeslice Pattern: Core Concept
Every Entity Has Validity Periods
The Nexus Model Architecture embeds time dimensions directly into every mutable business entity:
- ValidFromDate: When this version of the record became active
- ValidToDate: When this version was replaced (9999-12-31 for current)
- SourceSystemId: Tracks which system provided the data
| Column | Type | Purpose |
|---|---|---|
ProductId |
bigint IDENTITY |
Surrogate key (auto-generated) |
ProductNK |
nvarchar(100) |
Natural key from source system |
ProductName |
nvarchar(200) |
Current product name |
Price |
decimal(18,2) |
Current price |
ValidFromDate |
date |
When this version became active |
ValidToDate |
date |
When this version was replaced |
SourceSystemId |
bigint |
Which system provided this data |
Point-in-Time Queries Made Simple
Timeslices enable "as of" queries without complex joins:
--- What was the price of Product X on 2023-06-15?
SELECT ProductName, Price
FROM Product.Product
WHERE ProductNK = 'ProductX'
AND '2023-06-15' BETWEEN ValidFromDate AND ValidToDateBenefits: Why Timeslices Matter
Accurate Historical Reporting
Traditional approaches often approximate history. Timeslices provide exact historical accuracy:
- No approximations: Historical trends reflect reality, not calculations
- Regulatory compliance: "Show me what we knew about this customer on [date]"
- Audit readiness: Complete change history for every record
- Dispute resolution: Verify customer claims with historical data
Real-World Examples
A customer service dispute about a product price change? Query the exact price at the time of order—resolved in seconds, not hours of investigation.
Automatic ETL Updates
When source data changes, the framework automatically:
- Closes the current timeslice (sets ValidToDate to yesterday)
- Creates a new timeslice (ValidFromDate = today, ValidToDate = 9999-12-31)
- Preserves complete historical accuracy
Implementation: Automatic in NMA
The Nexus Model Architecture handles timeslice management automatically in ETL:
Changes"] --> B["ETL Detects
Changes"] B --> C["Close Current Timeslice
ValidToDate = Today-1"] C --> D["Create New Timeslice
ValidFromDate = Today"] D --> E["Complete Historical
Record"] style A fill:#e3f2fd,stroke:#1976d2 style E fill:#e8f5e9,stroke:#388e3c
Real-World Results
- Pricewise: All dimensions track historical changes automatically
- Creative Clicks: Order status history enables accurate revenue attribution
- Regulatory compliance: Complete audit trails out of the box
Timeslices: The Foundation of Trustworthy Analytics
In a world where business data changes constantly, timeslice-based history provides the foundation for trustworthy analytics. Every report, every dashboard, every decision can be traced back to what was known at the time—not approximations or calculations.
The Nexus Model Architecture makes this powerful capability automatic, requiring no special development or complex queries. Point-in-time accuracy becomes the default, not a luxury.
For technical implementation details, see our Resources page.
