The search for novel and efficient energy materials to enable a sustainable future is one of the most pressing challenges of our time. Translating materials from the laboratory to commercial application remains time-consuming, and the discovery of materials for technological applications has largely relied on heuristic, trial-and-error approaches. With my academic training in Physics, Chemistry, and the Computational modeling of materials, I aim to apply AI/ML techniques to bridge the gap between computational predictions and experimental realisation.
My graduate research focuses on developing digital tools to capture the effects of disorder and temperature in computational workflows, which are currently limited to ordered crystals when using first-principles methods. This enables a more systematic exploration of chemical and configurational space and yields actionable insights into composition–structure–property relationships, thereby accelerating the experimental discovery of functional materials.
My research projects span the pedagogical development of toy models and machine learning models for superconductors, density functional theory (DFT), and universal machine learning interatomic potentials (UMLIPs).
Large-scale text-mining pipeline extracting 35+ material properties from 4,000+ scientific articles to build a unified superconducting materials database supporting ML modelling and data-driven discovery.
Universal Machine Learning Interatomic Potentials applied to the vibrational physics of doped GeTe thermoelectric systems at a fraction of DFT computational cost.
First-principles investigation of Raman vibrational modes in B-CHap to support experimental measurements, identify characteristic vibrational signatures, and understand structure–spectroscopy relationships.
Machine learning models using Quantum Structure Diagram-based descriptors to predict superconducting critical temperatures, enabling accelerated data-driven materials discovery.
Pedagogical toy model demonstrating how two electrons with identical charge form Cooper pairs via effective attractive interaction mediated by lattice vibrations.