If you want to become better and better, you need one thing above all else: good data. However, it is often the case that the available information is hardly usable. It is our mission to improve the quality of our information and make it easier to analyze and thus more usable – for the well-being of our patients.
Based on a huge amount of data and using the latest technology such as artificial intelligence, machine learning or neuronal networks, we develop smart solutions that can respond intelligently. Our work involves more than just data analysis or identifying errors and correlations; we deliver results that benefit the entire company – whether in ideation, developing innovations or improving products and processes.
In this way, Data Scientists create added value for the whole company – as well as for our patients. Because by developing powerful solutions while keeping down costs, we enable more people around the world to benefit from our therapies. And we can support medical staff with solutions and products that are increasingly reliable and easier to use. That relieves the pressure on them in their day-to-day work and boosts patient safety.
At Fresenius, patient welfare takes top priority and governs everything we do. This is also our approach when it comes to improving our processes. We contribute to this with Data Science and Business Intelligence – collecting and preparing information and making it ready for analysis. In this way, we not only make treatment and everyday life more comfortable for patients, but at best, we also prolong their lives.
What is special about Data Scientists at Fresenius? They are people with a technical background of almost any kind or a degree that involves analysis and are interested in others’ opinions. If you are also creative, like to tackle difficult tasks and enjoy treading new paths, then we should talk. What’s more, you should be able to communicate and provide advice when dealing with different specialist departments, management and IT. An international outlook and diversity are equally important for our work. Different perspectives bring new impetus and lead to exciting approaches, which we discuss in a close, personal exchange – passionately yet objectively.
Power BI, Grafana
Azure Cloud Environment, Python
Jupyter Notebook, Kubernetes
TensorFlow, SciKit-Learn
Numpy, Keras
Denodo, Knime, Excel
We try to predict when a machine or parts of it will fail based on certain early warning indicators, such as error memory entries, warnings in program logs and sensors, anomalies in behavior, or noises. Based on these values, we can recognize whether a part has reached a certain life cycle and/or will fail at a predicted point in time. Technicians can then replace parts or prevent disruptions in good time.
Complaint Classification is designed to optimize processes used to deal with questions and complains from patients and customers. We filter requests using artificial intelligence that separates simple cases and requests from more complex ones so that they can be automatically forwarded to the appropriate teams for response.
When a fistula – an artificial link between an artery and a vein – gets blocked, it is essential to act quickly. In our E-Stethoscope project, we give medical staff an electronic stethoscope that recognizes such situations with the help of artificial intelligence, regardless of the nursing staff’s experience. That means more safety for patients.
During the COVID-19 pandemic, the structure of the supply and demand of drugs changed practically overnight: Non-essential operations were cancelled or delayed, while the number of patients on ventilators rose, and with it the need for certain anesthetics. Within a short period of time, we developed dashboards that give management a daily overview of inventories, production capacities as well as stocks currently in transit.
The aim of our Manufacturing Execution System is to abolish paper-based controls and documentation in production. Not only does this produce large amounts of paper, checking the documents by hand for quality assurance is also inefficient and prone to errors. This project involves digitizing the entire process, which allows us to make quality assurance more automatic, as deviations from the system are displayed.