A large central bank in Asia

The Challenge:

The client through its Data Warehousing project aims to make use of technologies such as relational and multidimensional database management systems, and meta data modelling and repositories so as to build an integrated database management system that provides decision-makers, analysts and researchers with online and real-time access to a central repository of clean and consistent historical and current data. Further, this project will use Data Warehousing technology to enable business users to use business language and terminology that they are comfortable with and are using in their daily tasks. The bank has a team of statisticians and economists who use historical as well as current data for the purpose of forecasting key macroeconomic variables. The project is expected to provide valuable support in their research and analysis work.

The prime objectives of this Data Warehousing project are:

	to develop an integrated repository of current and historical data, encompassing all operational and research areas
	to build a state-of-the-art decision support infrastructure with online analytical processing (OLAP) capabilities, which includes providing a multi-dimensional and subject-oriented view of the database
	to build a user group-specific data mart, based on the central repository
	to provide extensive metadata
	to enable Web-based access to the warehouse

The central bank chose i-flex to provide a comprehensive business and technology solution, covering all phases of this critical project. The project is not a substitute to the existing data collection and compilation process of any department but an attempt to gather the information scattered around the institution.

The project covered the entire functioning of the bank, including the nineteen departments at the central office and regional offices located in major cities. It aims to get all the data available in different parts of the organization to work together, and synthesize data into a modern database with a metadata "catalog". The bank has many reporting systems, wherein information flows from banks, corporate and other financial intermediaries at various frequencies. The departments within the bank are using different operating systems and platforms to process and analyze the data. The information resides in many different files and in database structures developed by different vendors. Implementing a Data Warehouse solution to source data from heterogeneous sources, clean, filter and transform the same and store it in a structure that is easy to access, and enabling business users to understand data in the terminology they are comfortable with, was a highly challenging task.

Solution offered:

i-flex addressed the requirements using a goal-based approach. The five basic goals of the central bank are:

Goal 1 - Issuing and managing currency

Goal 2 - Conduct of domestic monetary policy

Goal 3 - Exchange rate management

Goal 4 - Maintaining the systemic integrity of the banking system and other financial intermediaries

Goal 5 - Banker to the government and other banks

i-flex has completed its study of existing systems and the analytical requirements of business users from all departments. This will be used in building a comprehensive data model for the central repository and also dependent data marts around major subject areas of the bank.

The project is structured in three stages as follows:

 information management study

 data model and domain harmonization

 operation management

As part of the Information Management Study, the i-flex team completed the source system study and user requirement survey. Data on the source systems was gathered using structured questionnaires, interviews with users, and from existing documentation. Information about the existing system was documented using Data Flow Diagrams (DFD), Entity Relationship (ER) Diagrams and i-flex's proprietary tool for storing data inputs, outputs, transformation rules and business terms. This analytical tool helps in pinpointing the source of information across the departments, with linkages to systems, related input forms, output reports, amongst others.

User requirements were obtained through structured interviews and workshop sessions with users from each business user group. The goals, objectives and vision were identified in discussions with the bank's senior management. The existing business and business processes were studied and tools and methods used by business users were analyzed.

During the Data Model and Domain Harmonization stage, subject areas for immediate implementation are identified and a data warehousing solution is designed. The Business Solution design includes preparation of an enterprise-wide data model and definition of analytical domains and data quality and transformation rules. The source data is mapped to the data model for designing the extraction layer and the data staging area. The Business Metadata definition is prepared based on inputs received from business users. The Technical Metadata definition is prepared based on source extract definition.

During the Operation Management stage, the architecture, tools and technology are finalized. The hardware is sized and the software and networking requirements are finalized. The physical architecture for the proposed data warehouse is implemented. The source data is transferred from the legacy systems to the staging area. A security management framework is implemented. Contingency plans and backup plans are put in place and operational procedures are set up.