Implementation of a DW/BI project - by a green-horn designer!

Things are exciting in the world of DW and BI. Many of us are comfortable with the terminologies and concepts like dimensions, facts, star-schema, snow-flake schema, types and categories of facts and dimensions etc. While still learning DW/BI in much detail, some of us might be still tempted to get real and start designing and modeling a data warehouse.

This kind of curiosity leads us to think about things like DW/BI project phases and implementation that happens in a real world scenario. To understand this at a very high level, we will consider the case of an accounting  firm and its need for a business analytics capability. As a DW/BI solution expert our job is to gather requirement, analyze, design and implement a technical solution.

1. Requirement gathering and understanding the data: Business processes and data are centric to our project and solution. Hence first and foremost task is to know the business processes and understand the data. The processes and sources that create data, grain and flow of data are all important things to be understood and documented first. Diagram 1 shown below, represents the current data flow of the accounting company for which this DW/BI solution is being built. There are multiple sources, a path and grain which defines the data. Once this picture is captured the next job is to plan the way of project implementation.

Diagram 1

2. Project plan and phases: After gathering the  requirement and having a basic understanding of the data, one is in a good stead to know how the project can be implemented and propose the plan to the client. This accounting company has no database or data warehouse in place. Hence the project will start from understanding data, 'ETL'ing it and warehousing it. So creating ETL logics, staging database and data warehouse are all part of the project. After creating the data warehouse, data can be fed to an OLAP cube and ultimately be used for BI purposes in the form of reports, charts, scorecards, ad-hoc queries etc. Apart from the milestone events shown in Diagram 2, generally more than one sub-events will precede or follow the main events.  

Diagram 2

3. The outcome - DW/BI solution: Diagram 3 represents the high level diagram of the DW/BI solution that will provide business analytics capabilities to the accounting firm. It will be now able to fire "what-if" queries to the BI system and obtain answers. Decisions of the company and recommendations to the client will be based on the answers provided by the BI solution to the business users and decision makers of the accounting firm!

Diagram 3

The accounting firm is counting on us while making their decisions! As DW/BI solution experts, it will be our job to provide the correct and most accurate BI answers. To be successful in such a job we need more technical skills and expertise along with domain knowledge. Count on Kimball and Ross to provide us just that!

Inputs:

1) Typical data flow in an accounting firm - Professor Boulger, MIS Dept., Eller College of Management