erwin – Worldwide Leading Tool For Data Management

erwin has more than 25 years of enterprise architecture (EA) experience, helping organizations better understand, architect and transform the critical enterprise domains.
With erwin EA, you’ll know how business functions are aligned and structured, understand the information that powers business operations, and see where technology and innovation is lagging or can be used to drive competitive advantage.
erwin EA takes the guesswork out of business and IT transformation, and tames the inherent complexity of major IT initiatives like application rationalization, migrating to the cloud, and implementing Big Data.

Basic principles

• Data are resource
  • Data are a resource with the highest value for whole business system and have to be accordingly managed.
• Data are accessible
  • Users should have access to those data, that are necessary for them to do the tasks they have to do, regardless of organizational or functional distribution of data.
• Common Business Glossary and common data definitions
  • Data (meta-data) should be defined in a consistent way throughout whole organizations for all business functions. Definitions of data meanings should be understandable to all users, unambiguous and accessible to all users
• Data security
  • Data access should be properly secured so that un-authorized access should not happen.

Data architecture artifacts

• Business Glossary
• Conceptual models
• Key success factors
• Logical business models
• …

• Migration models
• ETL models
• Physical models.
• …

• DDL
• DB catalogue..

Data Models Architecture – example

Sharing information between different type of Users

Data Architecture of models for a DW development – example

Development process – example

Typical issues, reasons for integration requests

 

• Different locations, different operating systems
• Different  data bases
• Different applications
• Different data structures
• Multiple entries and multiple versions of same data, more versions of truth

TOO MANY data – TOO FEW relevant business INFORMATION

Challenges for integrations of different Data sources

How to ensure:

• Unique naming and definitions of business terms
• Harmonized security levels
• Harmonized „code lists“ and classifications
• Harmonized structures and data formats
• Appropriate Data Quality

for the whole organization

• Between ERP…CRM…BI and other Information Systems

MDM – “Master Data Management”

Methods, procedures and tools, for ensuring, that all applications, all business dashboards and reports for a specific organization, would be conceived on one, central version (“master”) of master business data

• Examples of master data
  • Business partners, products, articles, services….
  • Main code-lists
• Some of main reasons for poor quality of master data
  • No organization for Data Quality Governance
  • Separate maintenance of the same data in different applications
  • Non-consistent transferring, copying, amendments … of data between different applications

More versions of truth, which one is more reliable?

• Should be in-line with general Quality policy
  • General guidelines : TQM/ISO9000/SixSigma…
  • Quality policy for a specific organization….

Basic condition for Critical Success Factors monitoring

• Impact of Data Quality on the goals of businesses and success of performance
• Some of the basic criteria for Data Quality
• Correctness
• Completeness
• Consistency of data among various data sources (applications/DBs,  reports, documents,…)
• Consistency of data regarding time dimension
• Relevance of data for business needs
• Traceability between summary and transactional data
• Timeliness
• …

Iterative (spiral) approach

• Iterations for a limited selected range
• Transparency of development
• Users are not aware of all requirements in advance
• Speed of completion of all phases of each iteration (requirements-analysis-design-programming-testing-integration)
• Possibility of re-definition of requirements for the next iteration after the completion of an iteration
• Possibility of parallel activities of the different teams

Iterative (spiral) approach – key conditions for successful implementation

• Iterativity can also be a problem …
  … as each iteration focuses on a different area (of data)!
• Integration can be ensured only if all iterations originate from the same Data Model