Engineering systems, studying models, writing down what matters.

I’m a software and AI engineer working across production systems, machine learning, and research. My engineering work has included RAG systems, LLM evaluation and monitoring, developer tooling, real-time systems, observability, and cloud infrastructure.

My research interests centre on model reliability: how models express uncertainty, when they should abstain, and how fine-tuning and reinforcement learning methods affect that behaviour. Earlier work explored speech emotion recognition and resource-efficient hyperparameter optimisation for pretrained models.

I use this site to publish technical notes, research, and public case studies—especially work that benefits from a slower, more rigorous explanation than a short update allows.