|
In today's fast-moving business world where marketplace challenges
change abruptly, information technology holds the key to success. The
strategic mission of IT departments everywhere is to deliver technologies
that drive optimized enterprise performance. Your place at the center
of your organization's business
strategy depends on being able to deliver high-quality information
to the right people at the right time.
Knack is an established, independent provider of data warehousing strategy
development, design, implementation and troubleshooting services. Our
methodology for data warehousing projects includes the following six steps:
| Risk
Assessment. |
| Because a Data Warehouse
project integrates a variety of data sources across the enterprise,
it involves numerous stakeholders, work processes, IT resources, and
business requirements. Our first step is to interview stakeholders
to discover various risks to the enterprise: the feasibility of the
project, its sponsorship, the sponsor's motivations, and the culture's
need and ability to support change. |
 |
| Analysis
of Findings and Definition of Project Scope. |
| Our second step is to
analyze and present our findings. If the client believes the rewards
outweigh the risks, we draw up a preliminary project model. The model
details the project scope, the business case for the project, and
an estimate of the completed project's return on investment. |
|
| Test
of Project Scope. |
| With a preliminary model
of the overall project, we then return to the stakeholders and test
their business requirements and existing data sources against the
project scope. The result of this inquiry is a high level specification
that includes requirements for the technical architecture, dimensional
modeling, and end-user applications. |
|
| Design
of Technical Architecture. |
| There are three elements
to the design of the technical architecture. The first is to design
the environments of the data staging, Data Warehouse, and applications.
The second is to develop criteria for product selection (hardware,
database, data loading, and data access tools). And the third is to
select the products and tools and to install them. |
|
| Dimensional
Modeling. |
| The end step to installation
is to "put everything together," which begins with the design
of a star schema. A star schema is a standard technique for designing
the summary tables of a Data Warehouse. "Fact" tables are
joined to a larger number of independent "dimension" tables.
Then we design the data load process (including the data staging area
of the bus architecture). Finally, we create strategies that ensure
conformity of dimension tables (where they are shared by different
star schemas) as well as accommodate any slowly changing data in those
tables. |
|
| Creation
of the End-User Application Environment. |
| The final phase of any
Data Warehousing project is to design the applications that will be
used to access the warehouse. Based on our previous stakeholder interviews,
we will design report templates that will provide users a view of
the data defined by their parameters. The same report template will
be able to provide users with a large number of views of the data
as well. The applications and report templates can be delivered within
a web-based environment, a tool environment, or a custom application. |
|
