As a Machine Learning Engineer, I specialize in developing and deploying scalable, production-ready models, leveraging a technology stack that includes JAX, Flyte, Azure, Kubernetes, and Docker. My role encompasses end-to-end model lifecycle management — from training and optimization to deployment and monitoring. Working in this position, I drive advanced machine learning projects, I use JAX for high-performance numerical computing and leveraging Flyte to orchestrate complex workflows. Azure serves as my cloud platform, where I utilize Docker and Kubernetes to manage containerized applications, ensuring a scalable, fault-tolerant infrastructure. In this role, I collaborate closely with cross-functional teams to deliver innovative solutions, tackling real-world challenges through data-driven insights and cutting-edge technology. My work focuses on enhancing both model accuracy and operational efficiency to support business goals.

As a R&D Lead, I am responsible for evaluating, developing and testing new methods and algorithms for similation, optimization and model-based design of experiment with applications in the bio-pharmaceutical industry for improving drug production. The models are hybrid deep machine-learning and first-principle models. I analyze data and run simulations with our models to support decision making, ranging from media and feed optimization, design of experiment to gene knock-down/knock-out optimization. Other than researching and devoloping new methods for the aforementioned tasks, I also write code that it is used in our software product. I use mainly Python, together with some cutting-edge machine learning libraries like JAX and Equinox. Of course, I am proficient in version control system like git.

I was PhD student at the international Max Planck Research School. My research area was model based control and optimization of bioreactors using machine learning models under uncertainties.

My research focused on optimal control and model predictive control of bioreactors using machine learning and physics based models. In the past I also helped the Trajectory Drilling team of Baker Hughes Company to develop automation solution for trajectory drilling in the Oil & Gas industry.

I worked in the Advanced Battery Management System team. In particular I worked on state and parameter estimation in Lithium-ion batteries. My work consisted in elaborating, implementing and testing different algorithms for state and parameter estimation.

I worked at the Faculty of Electronic and Computer Science Engineering and Automation Engineering on a fast implementation of a Moving Horizon Estimation for linear systems. The algorithm allows to rapidly solve Moving Horizon Estimation problems