STATWORX meets DHBW – Data Science Real-World Use Cases
Everyone is talking about data science, but what is the best way to use it in companies? What do you have to consider when planning an AI project? What are the risks and what are the potential benefits? It is precisely these questions that students of Baden-Wuerttemberg Cooperative Universities as part of the guest lecture with STATWORX have discussed.
Aspects of the lecture
At the beginning of June, our COO Fabian Müller and data science consultant Paul Mora gave a lecture as part of the DHBW industrial engineering program. The focus of the lecture was to make students aware of which aspects need to be considered when planning and evaluating a data science project. In addition to the financial risks, the ethical issues of using artificial intelligence were also addressed explicitly.
Fabian Müller, COO at STATWORX, regularly gives lectures at colleges & universities to actively raise awareness about artificial intelligence.
One of our missions at STATWORX is to share our knowledge with society. Talks at colleges and universities are a great opportunity to sensitize tomorrow's generation to the benefits and risks of AI.
Hands-on case study
As a graded homework, the students then divided themselves into groups and evaluated a data science use case they had devised themselves within the framework of a company. The team of Christian Paul, Mark Kekel, Sebastian Schmidt and Moritz Brüggemann succeeded. As described in the following abstract, the team dedicated themselves to considering the use of data science in predicting customer orders.
Consultant Paul Mora explains the AI Project Canvas to DHBW students.
Abstract: Application of artificial intelligence in the context of a fictitious company
The present case study provides an overview of the possibilities of AI-driven problem solving based on the fictitious and up-and-coming winter sports equipment supplier company. Four different use cases that benefit from the use of AI were analyzed within a feasibility impact matrix and the concept of AI-driven after-sales management was prioritized.
With regard to after-sales management, no innovative methods of sales promotion have been developed to date. Only the sending of vouchers, four weeks after receipt of the order, is already used. However, this is not an adequate solution for long-term customer loyalty. With the help of concentrated discount or voucher campaigns, customers should be encouraged to buy the products again at the right time in the future. The right time, i.e. the due date when the customer's needs arise, should be determined continuously using AI. By using AI, management hopes to understand the customer journey and predict it in the future. The sales-increasing measure is based on the concept of descriptive decision theory developed by Daniel Kahnemann and Vernon L. Smith, which empirically represents how decisions are made in reality. Descriptive decision theory defines incentives at the right time to meet current needs/needs as a central aspect of a decision maker's decision-making.
The Data Sciences Model Canvas was chosen as a tool for structuring the implementation process of AI within the company. The present machine learning problem, which is used to predict future order dates for customers, is to be addressed with the help of so-called “supervised learning.” Overall, the algorithm tries to find a hypothesis that makes assumptions that are as accurate as possible, which is a regression problem under categorized. If done correctly, customers are encouraged to buy as soon as their needs arise with the help of concentrated discount campaigns. Among other things, this also makes it possible to retain hybrid customers whose demand behavior is variable but can be latently influenced. The use of an intelligent after-sales management system thus enables long-term market and customer orientation.
Interest sparked?
The fully prepared report and a short and concise management presentation can be downloaded below. The report shows how data science can be used effectively within a company to strengthen customer relationships and make more informed decisions. The report also presents three other potential uses of AI and weighs up its advantages and disadvantages through the AI Project Canvas.
