Learns gate placement and parameter initialization together; energy error down to 10⁻⁸.
About
I am a PhD candidate at the School of Computing Technologies, RMIT University, working on AI for Quantum and Quantum for RecSys. My work aims to get reliable results from limited quantum hardware: fewer qubits, fewer shots and fewer VQE calls.
News
- Converted from Master by Research to PhD at RMIT.
- FunnelAudit preprint on arXiv.
- PDQUBO published in ACM TORS.
- HamQASBench and DreamQAS preprints on arXiv.
- HyRLQAS preprint on arXiv.
- Q-EVT short paper at RecSys 2025.
Publications
My name is in bold. The full list is on Google Scholar.
AI for Quantum
Reinforcement learning for quantum architecture search.
HamQASBench: equally accurate circuits can use anywhere from 2 to 62 gates.
Models only the costly post-VQE feedback; 1.6–2.0× fewer real VQE calls.
Quantum for RecSys
Quantum optimisation for recommender systems on today’s quantum hardware.
First gate-based QPU study of recommender feature selection; predicts the shots needed to beat classical SA.
Recommendation with only five qubits, about 90% fewer than one qubit per tag.
Builds the QUBO from counterfactual effects of features and feature pairs on recommendation quality.
Other recommender systems work
Audits which retrieval routes are responsible for an item being shown or dropped.
Education and experience
-
Oct 2026 – May 2029 (expected)
Doctor of Philosophy School of Computing Technologies, RMIT University
-
May 2025 – Oct 2026
Master by Research School of Computing Technologies, RMIT University Full scholarship. Converted to PhD.
-
Jul 2024 – May 2025
Honours in Computer Science School of Computing Technologies, RMIT University Moved to Master by Research.
-
Oct 2023 – Jul 2024
Remote Research Intern RMIT University
-
Sep 2019 – Jul 2023
Bachelor of Science Information and Computing Science, College of Science, North China University of Science and Technology
Research assistant