Projects

A selection of research and software development projects I’ve worked on, bridging quantum chemistry, machine learning, and open-source science.

Predicting Quantum Energies with Deep Learning

Description: Designed a novel SU(3)-invariant descriptor for predicting electronic system energies using deep neural networks.
Impact: Introduced an innovative descriptor that significantly outperforms traditional quantum chemistry methods in energy prediction.
Skills: TensorFlow, Deep Learning, Theoretical Chemistry.
Paper

Detecting Hydrogen Bonds from Molecular Dynamics

Description: Built a robust statistical framework for identifying weak interactions directly from MD trajectory analysis.
Impact: Demonstrated that key chemical concepts like hydrogen bonding can be discovered automatically through data-driven analysis, without prior human intuition.
Skills: Python, GROMACS, High-Performance Computing, Statistical Analysis.
Paper

Deriving Covalent Radii from First-Principles Data

Description: Extracted covalent radii for 14 elements from a dataset of 26,050 small organic molecules.
Impact: Provided a new set of covalent radii derived directly from high-quality first-principles data, improving accuracy in molecular modeling.
Skills: Scientific Programming, Linear Algebra, Computational Chemistry
Paper

ModelHamiltonian: Electron Integral Generator

Description: Developed a flexible software package for computing 0-, 1-, and 2-electron integrals across various quantum models.
Skills: Software Engineering, API Design, CI/CD
Paper | GitHub | Docs

PyCI: A Python-Scriptable CI Engine

Description: Created a library for setting up and executing arbitrary determinant CI and FanCI calculations with Python interfaces.
Skills: C++, Electronic Structure Theory, API Design
Paper | GitHub | Docs

GBasis: A Python Library for Gaussian-Type Orbitals

Description: GBasis is an open-source Python library for efficiently evaluating and integrating Gaussian-type orbitals (GTOs). It’s aiming to make basis set manipulation more accessible and efficient without sacrificing performance.
Skills: Python, Computational Chemistry, High Performance Computing
GitHub