I’m an AI systems researcher and a PhD candidate in the Large-Scale Data & Systems (LSDS) group at Imperial College London, where I have been working under the supervision of Professor Peter Pietzuch since 2021. My research focuses on Machine Learning Systems (MLSys) for scaling large language models, spanning distributed training and inference, and I enjoy working across the stack, from Triton kernels through PyTorch to cluster orchestration.

My early work, Tenplex (SOSP'24), introduced dynamic parallelism that lets distributed deep learning jobs change their resources at runtime. My current research, Scepsy, targets high-performance serving of agentic LLM workflows.

In summer 2025, I was a Research Scientist Intern at Meta, where I modelled the serving of block-diffusion language models inside Meta’s internal performance simulator, showing that they add no extra latency over autoregressive models because attention is memory-bound.

I also collaborate with my second advisor, Mark van der Wilk, on Gaussian Processes (GP), where our research aims to improve benchmarking of GP approximations for comparability and enhance the performance of GP inference.

Prior to joining the LSDS group, I completed both my Bachelor of Science and Master of Science in Informatics at the Technical University of Munich.