henrique foerste

I’m a Data Scientist and Embedded ML Engineer with hands-on experience in AI/ML research, embedded systems, and data analysis. I like turning models into reliable, reproducible solutions—especially when they must work within real hardware and production constraints.

At Samsung Electronics, I developed an embedded ML model using TensorFlow Lite Micro to classify physical exercises from accelerometer data. I also improved Galaxy Watch heart rate zone detection by building and iterating with C/C++, Python, and Bash, while keeping the workflow collaborative and versioned using Git and DVC.

In my AI Research Intern role, I worked with large-scale Samsung customer data using Apache Zeppelin and PySpark, and I built an Android demo application integrated with Samsung Health to prototype health features. I enjoy bridging research and product by translating insights into prototypes that can be validated quickly.

My background includes NLP-focused research at the University of Campinas, where I helped develop a training program for Portuguese data annotation and fine-tuned the UBAR model on 5,000+ chatbot conversations to improve conversational performance. I’m now looking for full-time opportunities where I can apply my AI/ML and embedded expertise to impactful projects end to end.