Donovan is a NITheCS Postdoctoral Fellow working in the field of quantum machine learning (QML), which sits at the intersection of quantum computing and machine learning. His research interests include the application of quantum machine learning to tasks such as classification, clustering, and anomaly detection, particularly in the context of astrophysical and astronomical datasets. He also has a strong interest in theoretical quantum machine learning.
He began his physics journey in 2017 at the University of Pretoria (Tuks), where he majored in physics and minored in mathematics and radio astronomy. During his Honours year, he was introduced to quantum computing and completed an introductory project on Grover’s algorithm, which sparked his interest in quantum technology and led him to pursue postgraduate studies in the field.
This decision led him to the University of the Witwatersrand, where he joined the Wits Structured Light Laboratory under the supervision of Prof. Andrew Forbes. He submitted his Master’s thesis in early 2022, titled High-Dimensional Entanglement in the Spatial Basis. His research explored how entangled photon pairs could be experimentally generated and manipulated for applications such as quantum key distribution.
Wanting to return to a more computationally focused research topic, he decided to shift his focus entirely to quantum computing. He volunteered part-time at IBMQ Research South Africa, working on introductory quantum chemistry problems before ultimately committing to a full-time PhD in quantum machine learning under the supervision of Prof. Francesco Petruccione at Stellenbosch University.
Outside of research, Donovan is a passionate football fan and an amateur field hockey player. He also occasionally plays piano, squash, and chess. Donovan absolutely loves ice cream.
Publications:
Pulsar classification: comparing quantum convolutional neural networks and quantum support vector machines
Published in Quantum Machine Intelligence 6 September 2024. A comparative study comparing quantum convolutional neural networks and quantum-enhanced support vector machines in the context of pulsar binary classification.
Classical-quantum approach to image classification: Autoencoders and quantum SVMs
Published and featured in AVS Quantum Science 5 June 2025. Hybrid classical-quantum approach to image anomaly detection using classical autoencoders trained by image reconstruction for abstract, meaningful feature extraction. A quantum layer, consisting of quantum-enhanced one-class support vector machines are used as the final layer.
Quantum kernel and HHL-based support vector machines for multi-class classification
Comparison of a standard QSVM and an HHL-based LS-SVM (Gabriela Pinrheiro) for multi-class classification on reduced SDSS data. QSVM is implemented in a one-vs-rest scheme. This preprint is still in review.
Quantum Spectral Clustering: Comparing Parameterized and Neuromorphic Quantum Kernels
Comparison of a parameterized quantum kernel with a QLIF neuromorphic kernel (Dean Brand) and a classical RBF kernel for spectral clustering. The parameterized quantum kernel is implemented by trainable parameters in the angle encoding layers. This preprint is still in review.
Workshops and Conferences:
Quantinuum Quantum Computing Hackathon
In April 2023, Donovan was part of a team that traveled to the International Centre for Theoretical Physics (ICTP) in Trieste, Italy, to participate in a quantum computing hackathon hosted by Quantinuum Quantum Computing. Their project focused on predicting solar cell output based on two input features using Quantum Machine Learning techniques.
Horizon Quantum Computing Hackathon
Later in 2023, Donovan took part in another hackathon hosted by Horizon Quantum Computing in Bali, Indonesia. For this competition, his team worked on an introductory project in Quantum Error Correction.
International High-Performance Computing Summer School
In 2024, as part of a two-person team, Donovan attended the International High-Performance Computing Summer School (IHPCSS) in Kobe, Japan. This intensive week-long program covered topics in parallel computing, including OpenMP and MPI.
Quantum Techniques in Machine Learning
QTML is an annual conference dedicated to the intersection of quantum computing and machine learning.
Donovan presented posters at QTML 2024 (Melbourne) and QTML 2025 (Singapore):
- Pulsar Anomaly Detection: QOCSVM with CAE-Extracted Features
- Quantum Spectral Clustering: Comparing Parameterized and Neuromorphic Quantum Kernels
Donovan is also a member of the organizing committee for QTML 2026, which will be hosted in Stellenbosch in December 2026.
National Institute for Theoretical and Computational Sciences (NITheCS)
Donovan also assists within NITheCS with colloquia, seminars, workshops, and schools in various capacities.