Christopher Ehmann

Originally started working on Line Item detection via NN based Object Detection. We then moved to NLP based approaches for classification and extraction of information from insurance documents. Responsibilities include: * Taking customers requirements and framing it as a machine learning problem * Training and deploying neural network models * Executing and presenting results of PoC's * Writing, maintaining and dockerization of machine learning microservices

Master's Thesis on information transfer between neural networks in tensorflow. I used explanations generated from a trained teacher network to aid a student network during it's training process.

Here I did a small project on image segmentation using classical and neural network methods to create a new dataset from already existing images in the Zalando database by removing the background.

Worked on building a python backend/javascript frontend web tool, that allows system biologist to upload and test molecular pathways models, while visualizing flows over time in the molecular network.

Testing of Inte:Ligand, a molecular modeling software.