Junior Data Scientist, Data Scientist, software engineer

One of the largest publishers in Germany is in the process of migrating their multimedia resources (articles and images) to a new redaction system. The challenge is to transfer all data whilst maintaining the data consistency and integrity in the new environment. Also, due to the large size of the dataset, a process that runs robustly and reliably run over a time span of months has to be implemented and deployed. In addition to that, the newly integrated data has to be analyzed by AI services. Therefore, we provide a framework built with NiFi that continuously enriches the content with AI algorithms and ensures that data quality as the dataset is fed with new input.

Technologies: Java, Groovy, XSL, NiFi, ActiveMQ, Nexus, Gitflow

Conception, planning and development of a solution for optimizing the operation of a district heating network using physical simulation, demand prediction and data from temperature and pressure measurements. Currently the project has moved to the second stage, with the goal of modelling and minimizing heat losses in the network

Technologies: python, numpy, pandas, tespy

Given video sequences of patients performing facial movements as part of a medical diagnosis test, the system should differentiate between the following degrees of Stroke: No Stroke, Mild, Stroke, Severe as well score for facial palsy. The system was realized using neural networks, implemented with python and tensorflow.
Results: 99% overlap of the estimation by the developed system with the medical opinion after a computer tomography scan. 

Technologies: python, openCV, TensorFlow