Graduate of Imperial College London (MEng Aerospace Engineering) and former Software Engineer in the Advanced Concepts and Robotics department at Airbus Defence and Space.
My work centres on building high-performance systems at the intersection of machine learning, control, and scalable computation, with experience spanning robotics, distributed data systems, and GPU-accelerated simulation.
A C++ molecular dynamics simulation engineered for scalability and hardware efficiency.
- Object-oriented C++ design
- Cell list optimisation reducing naïve O(N²) complexity
- OpenMP shared-memory parallelism
- MPI distributed-memory scaling
- CUDA acceleration of force kernels
- 90%+ GPU runtime reduction
Focus: parallel scalability and hardware-aware optimisation.
Hypernetworks in Deep Reinforcement Learning for Complex Adaptive Systems Control
Research on parameter-conditioned policy architectures for nonlinear adaptive systems.
- Implemented hypernetwork-based policy models
- Benchmarked in MuJoCo environments
- Analysed robustness and generalisation trade-offs
- Evaluated conditioning strategies for adaptive control
Focus: interpretability, robustness, and scalable RL architectures.
Languages
Python • C++ • SQL
Parallel & Performance
CUDA • OpenMP • MPI
Data & Machine Learning
NumPy • pandas • scikit-learn • TensorFlow • Stable-Baselines3
Systems
Kafka-based streaming architectures
Event-driven design
API integration
Telemetry ingestion over TCP/UDP
- Measure before optimising
- Design for scalability from the outset
- Separate domain logic from parallel implementation
- Prefer clarity over premature abstraction
- Make performance explicit and reproducible