Data Science Carousel: Lots of "business value" with data!

On November 6th, no less than seven prestigious and very different practical cases have given the participants the necessary inspiration. The meeting gave the attendees, all also working at one of the members of the platform, a nice insight into the diverse possibilities with data science and digitalization. The lively discussions and the mutual exchange of experiences resulted in refreshing and renewed mindsets! 

After arriving at the Nutreco Global Hub in Amersfoort, CIO Henry van de Ven explained that Digital & Data is structurally a part of Nutreco's business strategy and also how they create value to it from an IT perspective. With the large amount of data – gathered throughout the chain – several Proofs of Concepts (POC's) are elaborated into an Minimum Viable Product (MVP). Overall goal of these initiatives for Nutreco is to ensure that they leave the right footprint for the next generation and strive for good farmer management, health management and nutrition management to optimally nourish the animal/fish without the use of, for example, antibiotics. Several initiatives within Nutreco were presented during the program by Amy de Bruin, Benny van Haandel, Erik Beckers and Noel Lopez Gonzaga. 

Data-driven at Royal FloraHolland
In recent years, Royal FloraHolland has taken rigorous steps to transform digitally. Meanwhile, they have grown to a data-driven organization, where new platforms are rolled out for growers and their customers. Smart applications have been thought out within different processes. Like the global trading platform Floriday and the logistical and commercial operation. Two showcases have been explained by Remco Wilting, Chapter Lead Data Science and Data Engineering. First, Floriday Insights: The Google Analytics for the horticulture sector. In this tool, with the help of current data, the actors in this market get real-time insights into the market developments, price formation, volumes, benchmarks, etc. The second showcase was about a tool for a product photo. Which gives the buyers a better idea about what exactly is being bought and sales can be optimized. With the help of a deep-learning application, the quality of the photos is automatically assessed and growers are given advice to improve them on different characteristics. 

Detection of insurance fraud with data science at VIVAT
Data scientist Feddo Kalsbeek, working at VIVAT Insurances, explained the steps that have been taken to allow more data driven decision-making to take place. They also want to act faster and better with data science to increasing individual needs of their customers. Practical example is the improvement of risk estimations and predictive analyses. The showcase that Feddo presented was about fraud detection, a perhaps unclear but efficient business case: no less than 10-12% of all insurance claims are fraudulent, with a significant amount of money that is therefore unjustified being paid. Very fascinating was their insight into how they use "surpervised machine learning" in different insurance policies and what potential unjustified claims or misleading policy requests should be subjected to further investigation.

Members only: Click here for the report of the theme session and the presentations form Nutreco Data Science Team & VIVAT by Feddo Kalsbeek & Bart Rentenaar.