Scientist turned AI builder working at the intersection of machine learning, scientific computing, and real-world systems.
I come from a background in astrophysics and Bayesian inference, where I developed high-performance computational methods to analyse cosmological data from space telescopes. My research contributed to the discovery of three exoplanets through large-scale C++ simulations.
Today I focus on building robust, efficient AI systems that translate scientific ideas into practical tools. My work has ranged from Bayesian deep learning for financial forecasting to multi-modal machine learning platforms for real estate analytics and underwriting.
I enjoy working with small, ambitious teams exploring new applications of AI and building systems from first principles.
⚡ Energy-efficient machine learning
🧠 Scientific machine learning and probabilistic models
🌍 AI for geospatial intelligence and remote sensing
⛏️ Mineral discovery using ML and satellite data
🔍 Efficient adaptations of foundation models
Languages
- Python
- C++
Machine Learning
- PyTorch
- Scikit-learn
- Probabilistic modelling
Scientific Computing
- High Performance Computing (HPC)
- Bayesian inference
- Numerical simulation
Domains
- Geospatial & remote sensing
- Financial modelling
- Real estate market intelligence
Machine Learning Leadership — Real Estate AI
Led development of a multi-modal ML platform combining tabular, text, and imagery data to forecast property markets and support underwriting decisions.
Entrepreneurship — AI in Finance
Co-founded a machine learning hedge fund through Entrepreneur First, developing Bayesian deep learning models for financial forecasting.
Scientific Research — Astrophysics
Developed Bayesian statistical methods and HPC software for analysing cosmological datasets from major space telescope missions.
🛰️ I once co-wrote a C++ simulation to model exoplanet orbits — and the work contributed to the discovery of three planets.
I'm always interested in discussing ideas at the intersection of:
- machine learning
- science-inspired AI
- geospatial intelligence
- efficient and data-lean AI systems
If you're building something interesting in this space, I'd love to hear about it.