What We Are Not

So many companies start by describing the services they provide. We are different and in order to prove that we will start by telling what we are not.

• We are not a large consulting firm but we are very effective
• We don't know everything about technology but we know where to find the answers
• We are not "heads" down coders that can't communicate with senior level managment. We can develop a solution while at the same time work closely with senior managment

Data Warehouse Development & Management

In its simplest form Data Warehousing is a process by which data is extracted, cleaned, filtered, transformed, summarized and prepared and presented in a manner understandable by a business. A data warehouse takes a vast amount of data that is not easily accessible by the average user and presents in a manner that can provide benefit to an organization.

In general a data warehouse should be:

• Designed to satisfy the needs of business users and not day-to-day operational applications
• A clean and consistent representation of a data store in a manner understood by business users
• A historical, current and summarized view of the operational data
• Easily accessible via user friendly interfaces and decision support tools
• Create a sense of ownership for the business users which in turn promotes clean data in operational systems

Building a data warehouse is a time consuming process but it can be accomplished in a cost effective manner. One should expect to spend the majority of the time gathering user requirements and analyzing and understanding the data in the operational systems. Without a thorough understanding of these two tasks a data warehouse implementation can quickly go from being cost effective to costly.

One needs to remember that a data warehouse is not a product but rather a solution to a business problem that is solved using multiple products.

Technical Architecture Design

Architecting and designing a data warehouse is crucial. Time spent now on a properly architected warehouse will be time saved in the future. Be sure to get your hands around the methodologies surrounding data warehousing as much has changed and is still changing.

Many people believe that a data warehouse is a copy of the source system just on a different hardware or database platform. Using this type of architecture is simply creating a reporting environment for users.

When architected properly a data warehouse is actually an entirely different implementation and design. The data warehouse will become a powerful tool that gives users access to THEIR data in an easily understood format. When a user can understand the structure and content of the data they become confident in the results and begin using it to make powerful business decisions.

When architecting your warehouse you will know that you are going down the right path when the name "Ralph Kimball" starts becoming familiar.

Source System Analysis

Source System Analysis begins after the project's business requirements have been gathered. Having the business requirements aid in developing the data warehouse model and in identifying which elements/attributes are needed from the source.

Source System Analysis helps identify where the required data resides and allows the data warehouse team to profile the data to identify inconsistencies in the source systems data. Inconsistencies could be foreign keys that are missing from a parent table or data values the fall outside of an acceptable range. Any inconsistencies should be reviewed with OLTP team and corrected.

Performing Source System Analysis is a time consuming process and is often overlooked until unit testing and acceptance testing is started. Source System Analysis is well worth effort because it allows the project team to identify inconsistencies before development begins.