Database Design Works
 Database Design for Analytics

Data Warehousing, Analytics  and Data Integration Design

Welcome


Welcome to Database Design Works, a division of Database Performance LLC http://www.databaseperformance.com

  We architect, design and implement databases targeted to support analytics usage by data analysts, data scientists and business users.  With over 30 years of experience in database, data warehousing and analytics we have an optimized approach that speeds development and provides flexibility while delivering a robust solution.

Our mission is to ensure that users can find, understand, interpret and take action on the information contained in their data.  Our data architecture for Analytics, Dimensional Normal Form, combines the best features of the Inmon and Kimball approaches in data warehouse design to enable an efficient, business-oriented enterprise approach providing flexibility, manageability and deployment speed to databases used for Analytics.

Our Template-Driven ETL approach speeds development while offering complete data validation and auditing to complement our Dimensional Normal Form data architecture.

We deliver design and implementation classes for Dimensional Normal Form and Template-Driven ETL internationally.

Data Warehousing success factors          

 The most Important success factors in data warehousing are

  • Business usage
  • Time to Value
  • Flexibility and speed of handling business requirement changes

Data Warehousing Best practices              

 Best practices in data warehousing to achieve success are

  • To promote business usage
    • Build incrementally providing business value every 30 to 90 days
    • Design to deliver usability
    • Design to deliver rapid and predictable performance
    • Provide extensive business definitions of the data and structures in the warehouse
    • Provide easy to use self service analysis and reporting tools geared to the need and capabilities of the business user
    • Provide the data validation processes required to meet data quality requirements
    • Provide extensive tracking and auditing of the data acquisition process and reconciliation back to the source system. If the business doesn’t trust the data they won’t use it
  • To implement quickly and efficiently
    • The architecture should be as streamlined as possible containing as few data hops and as few different structures of the same data as possible while still providing flexibility and maintainability and ensuring completeness of process
    • ETL processes should be as much metadata-driven as possible
    • Design templates should be created for ETL processes and followed by all developers