Agile Methods in BI and Data Warehouse Projects
More business value, less administrative overhead: Applying agile methods in BI projects ensures faster access to business-critical functions and makes for much happier customers.
The waterfall model vs. customer requirements
The demands customers place on their data warehouse and BI solutions – as well as the speed with which such solutions are developed – are growing at the pace of technological progress. They include:
- Faster provision of BI functions (and subfunctions)
- Ensuring data quality and stability the first time around
- Incorporating desired enhancements and other changes, even late in the development phase
- Reducing administrative effort to the essential minimum
- Turning around requests in short order to keep up with dynamic business environments
- Quickly accommodating organizations that want to know what their BI systems can do
The waterfall model, which has proven effective in implementing the majority of data warehouse and business intelligence projects over the past few decades, has begun to reach its limits in light of these demands. New, more agile methods are now required.
Philosophy and approach
At cundus, we apply agile, adapted methods throughout the entire process of developing and implementing business intelligence projects. Over the course of our past projects, we have formulated an approach capable of meeting our customers' demands in even the most complex undertakings.
Our goal is to optimize our service provision by keeping bureaucratic red tape to a minimum and focusing on the interaction of those involved in our customers' projects.
In doing so, we keep the pure conception phase short in order to come up with workable BI functionality as early on in the development process as possible; we can then constantly improve on this functionality in brief iterative cycles ("sprints") before delivering the final product. This also enables our project teams to make the resulting business utility available to our customers in very short order. In addition, this approach makes it possible to accommodate a variety of requirements and changes on short notice, which leads to a higher overall level of customer satisfaction. At cundus, we have adjusted our usual agile process – as well as the attendant roles and responsibilities – to the demands of professional business-intelligence development.
- Data modeling
- Data quality / automated testing
- Reference models / templates
Customer benefits
Our agile methods enable us to deliver business value to your organization more quickly, file projects down to essential levels of administration and documentation, and achieve increased flexibility in rendering our services. While our customers' organizations do have to adapt to these methods to a certain extent, the resulting change in mindset brings with it the following advantages:
- Project results offering a high level of utility are provided to the customer's organization early and often.
- Changes and additional requirements can be incorporated into projects whenever necessary.
- Requirements are fulfilled in short order following their specification, which helps in keeping pace with customers' actual current needs.
- Specialist users can quickly recognize the possibilities a modern BI platform offers and incorporate them into their requests.
- Rapid, cyclical test series ensure a significant increase in solution quality.
- The effort involved in documentation and project management is kept to a minimum.
