Targit Business Intelligence Suite

Introduction

As a Business Intelligence Specialist, I have for many years been
creating my own interfaces using a myriad of software and hardware applications
to deliver the easiest method of providing BI to the company’s decision
makers. One of my major concerns in this process has always been that any major
changes in the user’s requirements for data analysis usually required the
intervention of Information Technology specialists to write new queries or
change the method of delivery and that most organizations IT Departments do not
support this requirement. This has being improving steadily in recent years
with major improvements in the delivery of information to the user through SQL
Server/SharePoint/SSRS, Business Objects and Oracle/IBM Cube reporting and the
growth of the skill set within the database support areas of companies. However,
all of these methods still may require the assistance of the IT department either
to change the requirements of the report, or to investigate the underlying data
further to see if a new report is required or a drill down on the data reported
is required. In this article, I would like to give you an overview of my
experiences with TARGIT Business Intelligence Suite, a system that I believe
delivers functionality within an easy to use intuitive client along with an
easy to maintain and create data warehouse solution.

Discussion

The first step in any BI Development is to investigate the
available data across the company and then create a store or links to the
available information to be accessed by your BI client. The same methodology
applies for whatever industry your application is being designed and built for
– Sales, Supply, Finance or Health. The company is always going to require some
form of aggregated data to report its success or failure. This data is usually
created in a form of multi-dimensional cube using whatever technology the
company has available (Microsoft, ORACLE, IBM) but may also include external
data sources and data stored in a relational database. In a data warehouse
solution, where the data is stored in an OLAP cube, the dimensions, measures
and kpi’s stored are then interrogated by whatever BI system is put in place to
provide insight to the company. Unfortunately most IT departments do not have a
BI ‘expert’ on staff, the DBAs will understand about multi-dimensional data but
it will not be set up within the organization unless a BI project has been
completed or started. In my experience this is where a lot of organizations BI
projects fail – due to the added expense or complication of hiring staff to
complete the project in the first place. Having worked on several large
projects, I find that my time is split between management advice, creation of
the application and more frequently technology transfer to those permanent
members of staff who will carry the flag once I have left. This is not a
problem but it should always be factored into any project on which you may be asked
to consult. Once the data store has been created, the main project aim should
be a consolidated approach to report generation, avoiding if at all possible
the situation where as Eckerson stated in 2002 the ‘spreadmart’ becomes the way
forward within the organization allowing many departments to try and create
their own mini warehouses and BI systems to satisfy a particular need, utilizing
Excel to link to the Data Store.

A Solution

In my experiences of setting BI systems to work for international
clients, it has become apparent that the major driver for the client’s
management team is the speed at which results from the data can be delivered
once the data store has been completed. I believe that this speed of
implementation can define the success or failure of the overall project.

One method I have used to good effect is to implement a complete end-to-end
BI Model with TARGIT. As a consultant, I find that the flexibility of the
architecture model allows for most data sources found in IT Departments and the
speed of set up for a company using one of the industry standard epos systems
superb. For example I have completed the creation of the multi-dimensional data
store and install of a TARGIT system to a major sales company within 5 working
days, using the TARGIT Accelerator plug in for SQL which allows the cube to be
created automatically and then amended as required. The TARGIT Client is
installed on the user’s machine and once the TARGIT client is linked via the
ANT Server to the data cube the user is able to select from all the available
dimensions and measures to create reports, intelligent dashboards and
scorecards. The ANT server acts as a control layer between the Active X
components that underlie the client and the ADO layer connecting to the data
layer. The server is used to allow automation of report production, multi
lingual interpretation of the dimensions, measures and underlying data without
the requirement to change the language of the warehouse along with hosting each
clients stored reports and making global reports available to all clients.
Installing the server is a simple task and can be hosted on the same server as
the SSRS/SSAS install or on its own dedicated server as long as it can be
reached via TCP/IP from the clients.

The TARGIT client allows users to choose the type of report they
require, the display method, what data they want to see. I have found that one
of the most valuable parts of the client is the ‘intelligent analysis tool’,
which allows non-technical personnel the ability to select statements in plain language,
which are then converted via the client into reports based on the user’s
selections. As an example, a user might select to analyze the measure of sales
by the dimension of time selected by a particular region. This is all achieved
by following a controlled path in plain text language. Once selected, the
report is previewed in the lower screen of the client – it is quite amazing to
see the reaction on a users face when they have created their first report
using this method. For the more experienced user, the client also allows the
selection of dimensions and measures to create reports. In my experience a non-technical
user who has a background in sales finds the system very easy to use and can be
generating a report within minutes – this is the real benefit of this system.

Architecture

The TARGIT Architecture enables data to flow from either a
multi-dimensional or relational data source. The example below shows the
Architecture map for a Microsoft solution.

Conclusions

The TARGIT BI solution that I have used many times allows a
company to achieve a very quick turnaround from the decision to implement a BI
Solution to creating meaningful results; this in turn allows the company to
profit from those results. TARGIT is not by any means a complete solution for
all scenarios however, it will be interesting to see how the company moves
forward into the area of Cloud BI in the future. I have also used many of the
following systems (below) to link to OLAP and RDBMs databases and recover
information and I will continue as a specialist to deliver Business
Intelligence to a company using whatever methods are available at the time. However,
I believe that TARGIT is one of the most complete methods of delivery of this type
of data with a minimum of interaction required by specialists once the system
has been set up and installed, which in this cost effective economy must be a
plus:

• Jaspersoft
• Microsoft
  SSRS
• IDS
  Scheer
• SAP
• Arcplan
• Microstrategy
• Tibco
• Logic
  XML
• Panarama
  Software

Additional Resources

TDWI Reeling in Spreadmarts
The BI Survey
Targit Business Intelligence

»


See All Articles by Columnist
Peter Evans