I'm a software engineer and physics researcher at King's College London. My passion is building AI systems at the intersection of physics and computer science — probabilistic modeling, GPU acceleration, and edge AI.
Work
I build production systems where correctness and speed matter — from GPU research pipelines to models running on a watch.
- Summer Researcher & KURF Fellow · King's College London
- Physics-informed neural networks for carbon nanoclusters. +80% prediction accuracy, −60% simulation time. Jun 2025 — present.
- Software Engineer & CEO · Kannan Industrials
- Shipped iOS apps — 1minute DOEShelp, iPong, DabCounter. CoreML tuned for Apple Watch, 70% faster inference. 2021 — present.
- Research Intern · Kennedy Institute, Oxford
- CNN workflows for biomedical imaging with mixed-precision GPU training. 40% over baselines. Nov 2019.
- Data Science Intern · NHS Digital
- LSTM + Word2Vec clinical-note classification. ICD-9 coding accuracy 42% → 71%. Nov 2019.
Study
Physics for the models, computer science for the systems. I like working where the two overlap.
- BSc Physics · King's College London
- Alessandro de Vita Computational Physics Prize 2024–25. Computational physics, quantum mechanics, statistical mechanics. 2024 — 2027.
- MEng Computer Science · UCL
- Algorithms, machine learning, theory of computation, distributed systems. 2021 — 2023.
- A Levels · Royal Grammar School, Newcastle
- A*A*A* in Mathematics, Further Mathematics, and Physics. 2012 — 2019.
Stack
Tools I reach for. The common thread is making things fast and keeping them measurable.
- languages
- python · c/c++ · swift · java · haskell · bash
- ai / ml
- tensorflow · pytorch · jax · coreml · scikit-learn
- systems
- gcp · aws · docker · kubernetes · cuda · ci/cd
- research
- openmp · lammps · slurm · airss · graphql
Reach
Open to AI engineering, high-performance computing, and research roles. The fastest way to reach me is email.