Objectives:
The term Architecture refers to the art and science of construction. More generally, architecture refers to an organised arrangement of components meant for optimising the function, performance, feasibility, costs and aesthetics of an overall structure or system.
Data Architecture and Enterprise Architecture deal with complexity from two points of view:
- Quality-oriented: focus on improving execution within business and IT development cycles. Unless managed, architecture deteriorates and systems gradually become more complex and rigid, creating risks for the organisation. Uncontrolled data production, data copies and spaghetti relationships between interfaces make organisations less efficient and reduce the realiability of data.
- Innovation-oriented: focus on business and IT transformation to meet new expectations and opportunities. Promoting innovation through disruptive technologies and the use of data has become a role in the modern Enterprise Architect.
These two factors require separate approaches. The quality-oriented approach aligns with traditional Data Architecture work, in which architectural quality improvements are realised incrementally.
Activities carried out by our Team:
Establishing the Data Architecture pratice
Data Architecture must be an integral part of the organisation architecture. If an organisation architecture function does not exist, a Data Architecture Team can still be established. A framework relevant to the type of business should be chosen. Views and taxonomies of the framework must be useful in communicating to the various stakeholders.
In general, an Enterprise Data Architecture practice includes the following work flows, carried out in series or simultaneously:
- Strategy
- Acceptance and culture
- Organisation
- Working methods
- Results
The Enterprise Data Architecture also influences the limits of the scope of projects and system releases:
- Definition of project data requirements
- Review of project data studies
- Definition of the data lineage impact
- Checking data replication
- Implementation of Data Architecture standards
- Lead data technology and renewal decisions
Considering the specifications of existing Data Architecture
Every organisation has some form of documentation of its existing systems. These documents should be identified and evaluated for their accuracy, completeness and level of detail. If necessary, they should be updated to reflect the current state.
Developing a roadmap
If one were to create an organisation from scratch (free from dependency on existing processes), an ideal architecture would be one based exclusively on data needed to run the organisation; priorities would be set according to the business strategy and decision could be made without the contrastaints of the past. Very few organisations are in this situation. A roadmap provides a means to manage these dependencies and to make farsighted decisions. It also helps an organisation to identify trade-offs and to formulate a pragmatic plan in line with business needs and opportunities, with external requirements and available resources.
Managing business requirements within projects
The architecture must not be blocked by constraints at the time of tis development. Data models and other specifications that describe an organisation’s Data Architecture must be flexible enough to meet future requirements. A data model at architectural level must have an overall view of the organisation together with clear definitions that could be understood by the whole organisation.