What is Business Intelligence (BI)?

Put simply, Business Intelligence is:

"The art of turning data into information and information into knowledge"

In more detail, Business Intelligence is a broad group of applications and technologies to help enterprise users make better business decisions by gathering, storing, analysing and providing access to data.

BI applications include systems for decision support, query and reporting, on-line analytical processing (OLAP), statistical and 'what-if' analysis , forecasting, and data mining.

Background to Today's Problems

For the past decade banks have focused on progressively rebuilding operational systems, but most have problems with old and new systems that lack integration. They are now faced with unprecedented demands, both from management and regulators, not only to deliver more information, but to demonstrate tight control over internal processes.
This presents major challenges, with most of their vendors merely supplying bolt-on modules to existing applications and systems integrators delivering solutions which are very expensive and slow to implement.

As a result, many bank IT departments have become adept at extracting data from different sources and re-presenting the combined information using spreadsheets.

This has often led to "Excel hell", causing significant concerns about the integrity of data which is often used to make critical decisions. This approach also does not satisfy bank auditors and regulators as the audit trails are poor or nonexistent.

Granville's core products address the demand for sophisticated, real-time management information within the banking sector. This is needed not only for strategic purposes, but is required to comply with increasingly complex operational and financial regulation.

What are Today's Problems?

Many operational banking systems do not make it easy to report on operational data. Typical problems are:

  • Reporting is often not centrally planned, but developed ad hoc to meet individual department or staff needs
  • Many different tools are often used to extract data e.g. Access and Excel. For Midas clients, this also includes IBM Query, Midas/Q and Optical Reporting Facility (ORF)
  • In addition, there can be many one-off solutions for specific purposes e.g. reconciliation, regulatory and compliance reporting etc.
  • There are often multiple databases and multiple spreadsheets across all bank departments, maintained by many staff. These are often very complex with databases feeding spreadsheets, which then feed more spreadsheets etc. This can mean significant operational risk
  • Many different extracts are difficult to audit and difficult to maintain or change to meet new requirements
  • Banks need to keep staff skilled in all these different tools, solutions, databases and spreadsheets. This can be a significant cost and an operational risk as often such knowledge is only in one or two staff. When staff leave, often the knowledge goes with them
  • Much redundancy, with many reports produced but no longer used
  • Cost to ensure all these reports still work when source systems are upgraded or changed
  • Analysis is mainly done with Excel, leading to similar problems e.g. no audit trail, no "single version of the truth", many unmanaged end-user applications etc.

What's changed recently to make Business Intelligence more achievable and affordable?

  • Memory is plentiful and cheap - huge amounts of data can now be analysed in real-time
  • Data management software and tools are now much cheaper
  • Business Intelligence software is more sophisticated and usable
  • Intranets are in place for internal web-based distribution and analysis

So it is now possible to build an Information Layer at reasonable cost.
"BI for the masses" is now possible and Granville Associates can offer such a solution for banks.

What are the challenges when creating a Business Intelligence solution?

Operation data in a bank is huge, can be inconsistent across systems and difficult to access and understand.

Transforming this data into useful information is a complex and painstaking process. This means:

  • analysing business and data requirements
  • building a data model to support the business requirements
  • mapping source system data to the data model
  • writing code to read, transform and join the source data
  • building the processes to access the source data and load a data warehouse
  • building the processes to access the data warehouse to create an information layer and present this information to users.

How have Granville created their Solution?

Granville has spent considerable time meeting these challenges by performing the tasks above and engineering their solution. This meant:

  • First designing a generic banking data model
  • Then building the solution using the latest data warehousing technology together with the highest performing OLAP technology (see Architecture)

Many BI companies talk about what could be done with banking data but considerable work would then be required to perform detailed analysis and create a solution.

With Granville, this work has been done - a solution exists and is live at clients.