Adastra helped one of the Czech leading banks with optimizing of its operations and implementing the new Data Management concept

To establish Data Governance some issues associated with data element registration and the responsibility for them had to be solved; an overview of the complete process behavior in the area had to be defined while the data care organization structure had to be determined. From a Data Management point of view we clarified the complex concept of accessibility in all areas, set out priorities on where to focus our efforts and, crucially, sought and received support for these processes across the entire bank.
 
Implementing Data Governance as the basis for the desired level of Data Management for the customer involved working with the data to eliminate:

• Poor decision making.
• Repeated re-invention of previously used procedures.
• Duplicating activities used for the development and operation of data systems.
• Duplicating solutions in systems.
• Inaccessibility of information to authorized users at the requested time and providing the appropriate level of quality.
• Ineffective usage of resources spent on preventing problems where costs exceed the benefits sought by the prevention.
• Misunderstandings in communication during problem resolutions and creating systems.

If data belongs to everybody – and therefore to nobody – there can be no improvement of its quality.

Project consultants had the task of preparing a concept for organizing work in the Data Governance area. While working on this we relied on DAMA-DMBOK (DAMA – Data Management Book of Knowledge) and the DGI framework (Data Governance Institute). For processing of more specific principles and work methodologies within Data Governance, we created models based on the Adastra EAF GG (Enterprise Architecture Framework – Governance Generation) framework for modeling which had already been used by the bank for other projects.

An essential part of the project was developing the concept for the corporate dictionary with the option of using information from the model. We verified the prepared concept with ‘Proof-of-Concept’ while using a solution by Semanta with all available components including the Air Module.

One of the targets of the Data Governance Pilot project was the creation of a foundation for long-term Data Management which would ensure that:

• Basic terms and definitions associated with Data Management would be generally known and accepted in the entire bank.
• Responsible employees would be aware of activities and methods associated with Data Management.
• Assigned employees would know their responsibilities, rights and duties associated with Data Management, and recognize the options available to them when using resources and devices related to Data Management
• There would be sustainable means for exact and clear-cut data definitions and information in the entire bank (dictionaries, models, metadata).
• There would be unified rules for proposals and the development of applications and practices, and knowledge for working with data and technologies would be shared.
• Core banking data quality would be monitored and evaluated.
• Security policy for working with data (responsibilities, owners of data, rules) would be defined, implemented and controlled.
• There would be an exact model of core data, a transformation model and data flow in the entire bank. If reasonable, unified technological resources would be used for these models.
• Data Governance would be implemented – a method for organizationally securing Data Management.
• Metrics for monitoring Data Management contributions would be created. Data Management contributions would be evaluated and consistent improvements would be carried out accordingly.  

It is possible to state in retrospect that the initial project ideas concerning the contributions of the Data Governance implementation and the implementation of supporting tools, such as models, dictionaries, assigning evidence, governance etc., were correct, and this area brings significant benefits to the life of the company up to date. It has been proved that:

• Existence of the unified dictionary and administration of models make defining and analyzing requirements for development and system changes easier.
• The models enable more effective and precise analysis of the impacts of change. Moreover it helps to speed up data incident solutions.
• Integration costs decrease, as it is possible to define distributed data and their transformations more exactly, and prevent the creation of duplicated data and systems and indirect data transfers.
• Another contribution is the improvement of testing quality (as testing requests answers faster) and the speeding up of change managements due to the given availability and quality of metadata.
• Exact defining of competencies and processes ensures faster processing of business requirements.
• The unification of all technological resources and efforts within the entire bank decreases operational costs with regard to their maintenance.