The development process in knowledge management solutions usually follows the traditional project model, where the desired solution is defined at the beginning and is implemented in iterative work phases.
In software development, new software versions are often released, but the same practicalities that work in software development have not previously been used in data warehousing.
In his dissertation, Mikko Puonti focused on shortening the time needed to implement business intelligence solutions:
"My dissertation is very practical. In the research, we solved practical problems in industry and used the Design Science Research method to evaluate the solutions".
Implementation automation saves time and increases developers' satisfaction
An essential part of the research was to combine data from different systems into one centralised place – the data warehouse.
"The implementation of the data warehouse is the most time-consuming phase of the project. During the research, we implemented a tool for automating the implementation of the data warehouse. It had tremendous meaning. We reached a situation where the developer did the same thing in an hour, which previously took an entire working day", Puonti summarizes.
However, automation alone does not shorten implementation from months to days. The entire organization that uses the information management solution must be involved in the change.
The development process is different when reports are released often. Previously, it was enough that at the end of the project, the customer accepted the project and did acceptance testing. In a new way, testing must be included as one of the stages of the development work.
"It was amazing to see how much more satisfied the report users and their developers were, Puonti says with satisfaction.
Master of Science in engineering Mikko Puonti's doctoral thesis in the field of information technology, Continuous Delivery in Data Warehousing, will be publicly reviewed at the Tampere University's Faculty of Information Technology and Communication Sciences in Rakennustalo auditorium RG202, Korkeakoulunkatu 5on Friday 9 December 2022 at 12 o’clock. The opponent will be professor Pekka Abrahamsson from the University of Jyväskylä. The Custos is professor Hannu-Matti Järvinen from the Faculty of Information Technology and Communication Sciences.
Photo: Essi Peltonen