Machine Learning Engineering
SEW-Eurodrive
(Industrie und Maschinenbau, >10.000 Mitarbeiter)
Schlagwörter
Forschung
Machine Learning
Architektur
Computer Vision
Cloud Computing
Node.Js
Transformer
Cloud Platform
ReactJS
Large Language Models
Skills
Machine Learning Engineer & Data Scientist | Specialized in Computer Vision | AI & Cloud Solutions
Passionate about transforming ideas into intelligent, scalable solutions. With 3 years of experience, I bring deep expertise in Supervised & Reinforcement Learning, Imitation Learning, Knowledge Distillation, Transformer Architectures, and LLMs .
I’m driven by curiosity, open-mindedness, and a fresh research perspective—always eager to explore new technologies and push boundaries. I thrive at the intersection of research and real-world application, combining academic insight with hands-on industry experience.
Skilled in deploying solutions end-to-end using Node.js, React, Docker, and cloud platforms.
Let’s work together if you want to build something impactful
Passionate about transforming ideas into intelligent, scalable solutions. With 3 years of experience, I bring deep expertise in Supervised & Reinforcement Learning, Imitation Learning, Knowledge Distillation, Transformer Architectures, and LLMs .
I’m driven by curiosity, open-mindedness, and a fresh research perspective—always eager to explore new technologies and push boundaries. I thrive at the intersection of research and real-world application, combining academic insight with hands-on industry experience.
Skilled in deploying solutions end-to-end using Node.js, React, Docker, and cloud platforms.
Let’s work together if you want to build something impactful
Projekthistorie
04/2025
-
bis jetzt
Optimization of machine learning models, monitoring and ensuring deployment through CI/CD techniques, software quality assurance, as well as the design and development of end-to-end ML applications using Large Language Models (LLMs).
04/2024
-
bis jetzt
Conception, Development, and Deployment of Cloud-Based Machine Learning Models and Modern Web Applications
Robert Bosch GmbH
(Automobil und Fahrzeugbau, >10.000 Mitarbeiter)
In this role, I was responsible for the end-to-end process of designing, optimizing, and deploying machine learning models on the Azure ML cloud platform, subsequently offering them as a service. I also developed web applications using Node.js and React, managing MySQL databases in the process. My work included semantic segmentation of camera footage for driver assistance systems.
In addition, I took the lead in transforming an older Node.js HTML-based application into a fully Dockerized Node.js and React framework, seamlessly integrating the new machine learning service. This modernized website is now actively used by more than 100 users.
In addition, I took the lead in transforming an older Node.js HTML-based application into a fully Dockerized Node.js and React framework, seamlessly integrating the new machine learning service. This modernized website is now actively used by more than 100 users.
11/2023
-
03/2025
Development of Behavior Planning Algorithms and Contributions to Open-Source Julia Packages
Karlruhe Institute of Technology (KIT)
(Automobil und Fahrzeugbau, 5000-10.000 Mitarbeiter)
In a part-time role, I developed behavior planning algorithms using Julia under Linux, while actively contributing to open-source Julia packages. Furthermore, I collaborated on the publication of research results, presenting findings at conferences and in scientific papers. This position demanded close cooperation with multidisciplinary teams, as well as effective communication of complex technical concepts to both academic and industry audiences.
10/2022
-
03/2024
Development of a Video-Based Driver Assistance System Using Machine Learning, Cloud CI/CD, and On-Premises Node.js Deployment
Robert Bosch Gm,bH
(Automobil und Fahrzeugbau, >10.000 Mitarbeiter)
I was responsible for developing video-based driver assistance systems using machine learning methods in Python, leveraging Azure Cloud services such as Azure ML and CI/CD, as well as Docker. My role encompassed the end-to-end design, implementation, and maintenance of a CI/CD pipeline, which was realized with GitHub Actions.
As part of this initiative, I created a multimodal model that combined regression and classification approaches with a black box optimization model, utilizing PyTorch for model development. The training data was sourced from an internal MySQL database.
Upon completion, the model was integrated into an existing Node.js front-end application and deployed on-premises through CI/CD strategies. This comprehensive, automated pipeline—from data acquisition and model development to final deployment—ultimately saved the equivalent of two headcounts in a customer project, significantly reducing operational costs.
As part of this initiative, I created a multimodal model that combined regression and classification approaches with a black box optimization model, utilizing PyTorch for model development. The training data was sourced from an internal MySQL database.
Upon completion, the model was integrated into an existing Node.js front-end application and deployed on-premises through CI/CD strategies. This comprehensive, automated pipeline—from data acquisition and model development to final deployment—ultimately saved the equivalent of two headcounts in a customer project, significantly reducing operational costs.
Reisebereitschaft
In der Stadt Stutensee mit einem Radius von 50 km verfügbar
