Business Intelligence Portfolio

Business Intelligence Portfolio

Gary A. Thompson
thompson2286@gmail.com
(360) 241-3703

Table Of Contents
Project Overview
T-SQL Samples
SSIS – Integration Services
SSAS – Analysis Services
MDX Samples
KPI’s, Excel Services
SSRS – Reporting Services
PPS – Performance Point Services

Project Overview
Introduction: This presentation displays examples of my work (code samples and development work) using
the Microsoft Business Intelligence tools.

Tools used:
T-SQL 2005
MS SQL Server 2005
SQL Server Integration Services 2005 - SSIS
SQL Server Analysis Services 2005 - SSAS

Audience:
Business Executives
IT Managers
Solution Providers

Project Goals:
Design and build a relational database schema populated by data from three different data sources. With different data types involved,
performed Extract, Transform, and Load processes to prepare the data and load the tables. Each step was accompanied with a
confirming email upon the success or failure of the job run. This process used SSIS in the SQL Services Management Studio, and
T-SQL queries to extract the data.
Design and build an OLAP database and cube using the data from the relational database to allow for detailed data searches using
MDX queries. KPI’s allow a visual representation to be viewed showing a comparison between actual results and performance
goals.

T-SQL Samples
T-SQL provides a
means of extracting
data from the relational
database. Many of the
queries we
constructed were
designed to verify the
load process and to
ensure the accuracy of
the data within the
various tables.

T-SQL Samples

Narrowing the
selection criteria
through the use of
the logical “WHERE”
helped ease validation
of data by limiting the
output to a more
rigidly defined subset.

SSIS – SQL Server Integration Services 2005
(Documentation)

Ask most Developers and Analysts what their least favorite and yet
most important task is on a project and the answer is usually
documentation. Clear, concise, and accurate documentation is
critical to the functioning of a project and even more important to
maintaining systems on an ongoing basis.

Each package requires it’s own documentation about authorship,
purpose of the package, maintenance records, etc., and the overall
project documentation is put in place to give an overview of the
higher level requirements that are met and the roadmap of the
packages within the project.

Project: AllWorks, Inc. (SSISStudentProject)
Author: Gary A. Thompson
Created Date: Oct. 1, 2009
Modified Date:
Description: This project loads the AllWorksDBStudent database tables: Employees, EmployeeRates, County, Clients, ClientGroupings, Division,
ClientGroupingsXClients, JobMaster, and JobTimesheets.
The project involves the following packages:
EmployeeMasterPackage: Loads Employee table from Employees.xls
EmployeeRatesPackage: Loads EmployeeRates table from Employees.xls
based on matching Employee table on EmployeePK.
ClientMasterPackage: Loads Clients table from ClientGeographies.xls
County table must load first, and be validated prior
to loading Clients.
SpecialGroupingsPackage: Loads ClientGroupings table from ClientGeographies.xls
based on matching ClientPK in Clients table. Data
is aggregated by grouping #.
DivisionDefinitionsPackage: Loads Division table from ClientGeographies.xls
ClientGroupingsXrefPackage: Loads ClientGroupingsXClients table from ClientGeographies.xls
based on matching ClientPK in Clientstaable, and
ClientGroupingsPK in ClientGroupings table.
JobMasterPackage: Loads JobMaster table from ProjectMaster.xls
JobTimesheetsPackage: Loads JobTimesheets table from ProjectMaster.xls
based on matching JobMasterPK in JobMaster table
and validating that the job is not closed prior to the
workdate being submitted.

These packages are included in the project (SSISStudentProject) and run from MasterPackage. Also included in the Master are four Database maintenance tasks: 1. Shrink
the database. 2. Rebuild the DB indexes. 3. Update the database statistics. 4. Backup the database. Upon the successful completion of these tasks the MasterPackage
processes through the above packages to load the tables. If the Database maintenance tasks fail, and email is sent indicating which task failed, and processing stops.
Upon the completion of each package contained in the MasterPackage, an email is sent indicating the success or failure of the individual package run. Success emails
contain a count of rows being inserted, changed, and invalid rows (where indicated in requirements). Invalid rows are written to a text file for review by the users.

The Control Flow provides a
roadmap of the processes
within the package, and also
provides a step by step
explanation of what the
various processes will
accomplish.

This Control Flow shows
the steps involved in a
backup and maintenance
process. Each step is
dependent upon the
successful completion of the
prior step.

The arrows here represent
the path that the data takes
and the order of execution
of each process.


Through the use
of script tasks we
are able to
extend the
functionality
Provided within
SSIS to include
customized
processes that
we define using
Visual Basic for
Access.

The Data Flow
Task provides
tools for
extracting,
transforming, and
loading data from
various data
sources. The
Conditional Split
shown here
directed the flow
of data into
different output
files or databases
based on
specified criteria.


SSIS provides a
graphical interface
to help give a
visual
representation of
the steps being
taken. Each symbol
represents a data
source, a
transformation
process, or an
output destination.
The arrows are
data pipelines that
show the flow of
data from one
point to another.


Containers allow
for the use of
iterative
processes. In this
case, when this
process
completes an
email indicting
the success of
failure of the
process is sent to
the defined user
group.


I’ve used a legend
to identify
abbreviations which
identify the data as
it moves through
the various steps.
Consistency keeps
errors at a
minimum and
allows for ease in
debugging when
things go wrong.


Using annotation
to label the
components and
data flow assists
in giving a quick
idea of what
takes place,
without going
into great detail.

SSAS – SQL Server Analysis Services 2005

Good design
leads to a well
functioning cube
structure.
Verifying the
relationships,
ensuring that the
naming will
display properly
for the end user,
and referencing
keys is all part of
good design.


The structure
of this cube
contains four
fact tables and
nine total
dimensions.


Relationships are verified in the Cube Structure area and
confirmed in the Dimension Usage area.


Calculations are
added in the SSIS
interface in much
the same manner as
they Calculated
Members are built
in MDX.

This calculation
allows for a default
value to be inserted
in the event of a
null value.


Building calculations
help in the readability
of KPI’s and allows for
their reuse in other
areas.

Additionally, modularity
of code makes for ease
of maintenance.


KPI’s (Key Performance Indicators) are designed to allow the end user
a quick graphical representation of how their area is measuring against
their goals. They also can be used for forecasting based on trend
analysis.

The databases are partitioned at 50% in a MOLAP format
(Multi-Dimensional OLAP) and are divided by year.

MDX: Multi-Dimensional Queries
Multi-

MDX gives us a
means for querying a
cube structure,
grabbing data from
various dimension
tables to use in slicing
the facts.

This query
determines the labor
rate for employees
on a project based on
their hours worked
and the total labor
for the project.

Multi-

Another query uses
the Generate
command to create a
subset of data from
different sources,
then uses this newly
defined subset to
organize the data for
output.

Multi-

Using a calculated member in this query provides formatting for
data fields that contain null values.

Multi-

Although this
Query itself
was not long, It
required
documentation
to provide
others with a
clear idea of its
purpose and
function.

Multi-

The use of
multiple
calculated
members
expanded the
functionality of
the query,
while
simplifying the
code for
greater
readability.

KPI’s – Key Performance Indicators

For this project, I
developed several
KPI’s to give the
business users a
quick visual
summary of the
performance of
selected measures
against a particular
goal. While KPI’s
can also be used
for forecasting, job
requirements did
not call for such
measurements.


This KPI show the increase in overhead expenses against a goal of 15%
being acceptable. The goals are defined as good, marginal, and unacceptable,
using a traffic light symbol to indicate the status. Green is good, yellow is
marginal, and red is unacceptable.


The KPI’s are
displayed here in
Microsoft Excel.
This allows users to
see performance
measures without
requiring any special
installation of
software for
viewing. I created a
pivot table to hold
the data and KPI
information.


The beauty of using KPI’s in pivot tables to display progress is that
the measures can be adjusted on the fly. This allows the user to select
a different way to slice the data to meet their needs.

SSRS- SQL Server Reporting Services 2005
Reports are built to
display information
about how a
company is
performing with
regard to specific
goals. In this example,
if the percentage of
increased overhead
expense exceeds a
certain level, the
value is displayed in
red to supply a quick
visual representation
of company
expenditures.

SQL Server Reporting Services 2008
This is a
parameter driven
report that
provides a look at
several previous
months
production.
Depending on
selection criteria,
the header
information
changes to match
the selections
based on using
report items
within the body of
the tablix.
Footer
information is
developed using
calculations
based on report
items in the body
of the report.

SQL Server Reporting Services 2008

This report runs on a group
level and is navigated using
a document map on the top
level group to create a table
of contents in the PDF file.

Each item in the table is a
separate report based on
the groupings built in the
master report. The charts
are incorporated in the tablix
header so that they
consistentl ymatch the data
in the tablix with regard to
selection group.

Shading of alternate lines is
based on the producer level
changing, not simply every
other line in a different color.

PPS- Performance Point Services 2007

Graphs and charts come in a variety of shapes and sizes to
accommodate all types of reports. Line charts give a comparison
of two or more measures in an easy to read manner.


Different types of charts are
combined to display
different values being
tracked. This chart
represents the percentage
of profit for an organization
and the performance
broken down by county.


Scorecards and dashboards
can be displayed in many
fashions. One popular style of
dashboard displays as a traffic
signal, with green
representing progress ahead
of goal, yellow being a
marginal area near goal, and
red indicating being below
target.

Business Intelligence Portfolio

More Related Content

What's hot (20)

Viewers also liked (7)

Similar to Business Intelligence Portfolio (20)

Business Intelligence Portfolio