Data Scientist at Materna SE

During my time at Materna SE, a large IT company with over 4,000 employees, I honed my expertise in data science while working on projects for both the public sector and various industries. My role involved consulting with customers to understand their needs and provide tailored solutions. A significant part of my work was developing AI systems for tasks such as predictive maintenance, object detection and recognition, and sentiment analysis, which helped improve efficiency and decision-making for our clients. In addition to this, I played a key role in designing and implementing Question & Answering systems that streamlined information retrieval and improved user interaction. A precursor to the modern RAG-Systems (retrieval augmented generation).

In some cases, when needed, I also generated synthetic data, including 2D and 3D images, as well as simulations for predictive maintenance. These data sets were crucial for training robust models, especially in scenarios where real-world data was scarce or difficult to obtain.

As I learned and understood early on that every AI project is only successful with the complete deployment of the developed models, I decided to focus on MLOps to support the deployment and scaling of our machine learning models. With that I set up a comprehensive MLOps architectures, using ClearML, as well as mlflow.

My work also included creating data visualizations, which were instrumental in communicating insights and results to stakeholders in a clear and compelling manner. I worked with tools such as Grafana or Prometheus, but also developed customized dashboards with Plotly Dash.

Lots of my work was often done in a team effort utilizing software development frameworks like Scrum or Kanban, supported by often used Software like Confluence or Jira.

Beyond my technical responsibilities, I took the initiative, together with two co-workers, to organize a Data Science Community within the company, fostering collaboration, knowledge sharing, and innovation among data science professionals and like-minded people. I also shared my expertise with a broader audience by giving a talk at one of Materna’s developer summits, where I discussed the latest trends and techniques in MLOps. In order to grow the department I was also often invited to support in finding new coworkers by accompanying interviews. These experiences have not only sharpened my technical skills but have also allowed me to contribute to the growth and success of the company in a meaningful way.

During my four and half years at Materna my skills and worked helped develop a new AI and Data Analytics department into a striving and integral part of Materna.