Levin Kobelke Email & Phone Number

Mannheim, Bw, De

Ludwigshafen Am Rhein, Rheinland-Pfalz, Deutschland

As a data scientist at TrinamiX, I take responsibility for the development of machine learning models for NIR spectroscopy data analysis. In this role, I work closely with cross-functional teams to design and implement novel approaches for spectral data preprocessing, feature extraction, and classification. I also collaborate with experimentalists and customers to design and execute experiments to generate spectroscopic data and validate models. Responsibilities:✓ I analyze and interpreted spectroscopic data using statistical and machine learning tools, and present results to stakeholders.✓ I develop and maintain code repositories, databases, and documentation for data analysis workflows, and contribute to the development of TrinamiX's software infrastructure.✓ Collaborating with software engineers to integrate data analysis solutions into TrinamiX's software products and participating in agile development processes and code reviews.✓ Implementing automated testing, continuous integration, and continuous delivery (CI/CD) pipelines for data analysis workflows and software products, using tools such as GitLab, AzureDevOps, and Docker. ✓ I develop and maintain DevOps and infrastructure management tasks, including cloud deployment, monitoring, and troubleshooting.

Heidelberg, Baden-Württemberg, Deutschland

As a master's student at the DKFZ, I led a machine learning project focused on medical image classification and classification failure detection. In this role, I developed and implemented SOTA approaches for image preprocessing, feature extraction, and classification, using deep learning techniques such as convolutional neural networks (CNNs). I also designed and trained models to detect classification failures, which is an important consideration in medical imaging applications where misclassification can have serious consequences. Responsibilities:✓ Developed and implemented machine learning models for medical image classification, using techniques such as CNNs, transfer learning, and data augmentation.✓ Optimized image preprocessing pipelines for medical images, including normalization, rescaling, and cropping.✓ Designed and trained models to detect classification failures, using techniques such as outlier detection and uncertainty quantification.✓ Analyzed and interpreted classification results, and presentation of the results.✓ Conducted experiments to evaluate the performance of image analysis models, and developed methods for validating model accuracy and robustness.

Master Of Science - Ms, Biochemistry, 1.1

This program allowed me to build a solid foundation in the principles of biochemistry, informatics and data science. I learned how.

Bachelor Of Applied Science - Basc, Biochemistry, 1.2

Activities and Societies: Mentoring of new Students Tutoring in Maths, Physics, and Organic ChemistryThis program teaches a solid.