I'm Boris, a Computer Science & AI student at IE University. I primarily have a passion for backend engineering, particulary at the intersection with machine learning, although in general I just love building. Competitive athletics have shaped my approach to engineering: discipline, consistency, and long-term improvement. These are qualities I try to apply in software development wherever possible. I'm Dutch born, grew up in the United States and am currently studying in Madrid.
My tech stack: Python, JavaScript, Java, C, Bash, Pandas, pytest, Apache Kafka, MLflow, Grafana, Prometheus, Redis, Azure
Distributed Inference: An experiment in multi-node LLM inference using PyTorch and DeepSpeed. Our goal was to run and benchmark the OpenLLaMA 3B v2 model across GPU nodes under Slurm, exploring pipeline parallelism, scaling behavior, and performance tradeoffs. The final distributed run wasn’t achievable given the resources available on the cluster, but I found the project super interesting and learned a lot about how large language models work at scale.
Mortgage Predict: A mortgage analytics project exploring classification and regression models using scikit-learn, XGBoost, and LightGBM. A major focus was merging and validating multiple real-world datasets, followed by extensive data cleaning, feature selection, and model comparison with SHAP.
YAYA: A backend-first event planning app built as a microservices playground. I used FastAPI, gRPC, and Celery to separate reads, writes, and background work so user-facing requests stay fast while metrics and other non-critical updates run asynchronously. The project focuses on clean service boundaries, practical CQRS-style patterns, and production-inspired backend design.
Email: [email protected] · Resume · LinkedIn