Zürich Metropolitan Area
Defining the intelligence layer and core abstractions for industrial AI systems, enabling ML and LLM-based reasoning over sensor, geospatial, and operational data.
Worked across international roles in AI, data, and software engineering in Sweden, the US, and Japan. C++ Software Engineering (Gothenburg): Improved radar sensor data pipelines with fault detection and stability enhancements for safety-critical SPA2 vehicle systems. Data Engineering (Mahwah, USA): Led migration of production pipelines from AWS to Azure, integrated Snowflake, and introduced Airflow for scalable orchestration and ETL performance gains. Financial Data Science (Gothenburg): Built time-series forecasting models (Prophet) for operational cost planning used in executive decision-making. Generative AI Engineering (Tokyo): Fine-tuned LLaMA-2 with LoRA and DPO for culturally adapted multilingual sentiment prediction across APEC markets; scaled training with Accelerate and FSDP.
Deep Learning Engineer at Zenseact (Gothenburg), developing vision-based ego-motion prediction, training deep learning models for object detection, and integrating multimodal sensor data for high-precision localization, contributing to Volvo Cars’ autonomous driving sensor strategy.
Developed ROS-based software for a student robotics project; integrated control of a Ridgeback mobile base, Doosan robotic arm, and gripper, providing a unified starting point for path-planning assignments
Teaching assistant for the Machine Perception (RO47004, 5 ECTS), Deep Learning (CS4240, 5 ECTS) and other master-level courses. Tasks included updating course content prior to the start, supervising group assignments throughout and grading.