Leroy Freyler

Machine Learning Researcher with a solid academic foundation in Computer Vision, Inverse Problems, and Neural Networks, complemented by hands-on experience developing ML models and data pipelines. Proficient in applying advanced ML techniques to real-world challenges, with an academic focus on biomedical data processing

MLOps Engineer (Working Student)
SensXPERT, Munich, Germany | October 2023 – Present
As an MLOps Engineer at SensXPERT, I optimized cloud computing setup time by 30% using Azure, Databricks, Docker, and Spark, which led to faster deployment of machine learning models. I developed and maintained MLOps pipelines, seamlessly integrating PyTorch models with Python and Spark to ensure efficient data processing and model deployment. My role involved collaborating with an international team, ensuring the smooth integration of machine learning solutions across various cloud platforms.

Data Scientist in Predevelopment (Working Student)
BSH Hausgeräte GmbH, Dillingen, Germany | February 2021 – July 2023
In my role as a Data Scientist, I worked on a dataset augmentation project that enhanced data diversity for 2D, 3D, and audio-based data using PyTorch, OpenCV, and TensorFlow. I developed predictive models for noise emission based on 2D and 3D image data, which significantly improved the accuracy of product development processes. My work required collaboration with multiple departments and external companies in an international environment, driving innovation and integrating advanced computer vision solutions.

Assistant Scientist
Fraunhofer IOSB, Karlsruhe, Germany | December 2019 – December 2020
At Fraunhofer IOSB, I developed and implemented real-time skeleton estimation algorithms for human posture detection using Python, PyTorch, and TensorFlow. Additionally, I engineered anomaly detection systems for human movement patterns, contributing to enhanced public safety measures through advanced neural network models. I also utilized TensorBoard to track experiments and optimize deep learning model performance.

Technical Trainer
Hekatron Technik GmbH, Sulzburg, Germany | March 2015 – August 2017
As a Technical Trainer, I designed and delivered comprehensive training programs for technical apprenticeships and study courses. I developed both theoretical and practical curriculum content aligned with industry standards and emerging technologies. I managed multiple training projects, ensuring their timely completion and maintaining high-quality outcomes. Additionally, I mentored students, guiding their technical skills and professional development.

M.Sc. Electrical Engineering and Information Technology
Technical University of Munich (TUM), Munich, Germany
During my Master’s program at TUM, I completed coursework that provided a strong foundation in both electrical engineering and biomedical applications. In the Practical Course Neurosignals, I focused on measuring and analyzing neuroelectrical signals, including audio-signal processing and multi-channel EEG measurement. I also studied Biomolecular Electronics, where I delved into electronic structures, biomolecular surface functionalization, and electronic biosensors, with a focus on biomedical engineering.

My studies also included Introduction to Bioengineering, which covered the structure and function of cells and tissues, as well as advanced techniques in tissue engineering and biochips. In the course on Neuroprosthetics, I gained comprehensive knowledge about neuroimplants, electrical modeling of nerve cells, and the simulation of neural excitation and coding strategies for cochlear implants.

Additionally, I explored Graph Information Processing, which involved probabilistic graphical models, graph signal processing, and graph neural networks, with practical applications in clustering, classification, and network topology identification. My coursework in Solving Inverse Problems with Deep Learning focused on applying deep learning techniques to inverse problems, such as image recovery, particularly in medical imaging.

Furthermore, I studied Signal Processing and Machine Learning, integrating these techniques with a focus on multidimensional digital signal processing (DSP), spatio-temporal signal sampling, and solving inverse problems in media processing. In the course on Biologically-Inspired Learning for Humanoid Robots, I investigated brain-inspired algorithms for motor control and learning in robots, including reinforcement learning and self-organizing maps.

B.Sc. Electrical Engineering and Information Technology
Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany | 2017 – 2021
During my Bachelor’s program at KIT, I gained a comprehensive understanding of electrical engineering principles, which laid the groundwork for my advanced studies.

Bachelor Professional of Electrical Technology and Management
Chamber of Industry and Commerce (IHK), Freiburg im Breisgau, Germany | 2015 – 2017
This program provided me with a solid foundation in electrical engineering, coupled with essential skills in personnel management, business administration, and understanding corporate structures. This education also helped me develop leadership skills and gain practical insights into the industry.

Electronics Technician for Devices and Systems (Apprenticeship)
Hekatron Technik GmbH, Sulzburg, Germany | 2011 – 2015
During my apprenticeship at Hekatron Technik GmbH, I mastered circuit diagram planning and implementation, acquired proficiency in microcontroller programming, and gained hands-on experience in product development and quality assurance.