PhD in Computer Science (Lund University), currently Tech Lead at Theca.com. I work on ML systems, constraint optimization, and data-driven software infrastructure.
-
COMBO (Constraint-Oriented Multi-Variate Bandit Optimization): a toolkit for online optimization of LLM prompts or software parameters using live data. It supports formal constraints like token budgets and user tiers making it suitable for handling scalibility challenges in agentic AI. The system is built around Bayesian optimization and contextual bandits, deployable in Kubernetes.
👉 If you're working with LLMs or A/B tests, give COMBO a ⭐ on GitHub!
-
Base system open education labs at Lund University. The course focuses on system design and modern software practices at scale through web development.
-
FACE Theory & A/B testing My PhD work introduced the FACE model for understanding organizational and technical barriers to continuous experimentation. I also proposed practical frameworks to support controlled deployment and live testing in industrial software settings.