CapOpt: Capability-aware Superoptimization for Secure and Provably Faster Code
2026 International Conference on Object-Oriented Programming Systems, Languages, and Applications (OOPSLA) (CCF A)
My research focuses on building secure and efficient computer systems for AI infrastructure. I investigate Capability Hardware Enhanced RISC Instructions (CHERI) and complementary security mechanisms to create safe, isolated execution environments for AI agents. Additionally, I optimize systems software for AI training and inference to maximize performance and resource efficiency. My broader interests encompass cloud computing, cluster scheduling, and distributed resource management.
I am a Research Fellow at the University of Leeds and a Visiting RF at the University of Glasgow.
I am a co-founder of ACE3 AI LTD and a technical consultant on AI infrastructure at Kuaishou Technology.
I have served as a reviewer for IEEE Transactions on Parallel and Distributed Systems, a member of the EuroSys Shadow Program Committee, and Publicity Co-Chair for IEEE CISOSE JCC.
Selected publications
Google ScholarSelected work in memory safety, AI systems, and cluster computing.
SynergyScale: Optimizing Offloading and Task Partitioning for Efficient Model Training
2026 IEEE Transactions on Parallel and Distributed Systems (TPDS) (CCF A)
Sweet or Sour CHERI: Performance Characterization of the Arm Morello Platform
2025 IEEE International Symposium on Workload Characterization (IISWC)
StrongHold: Fast and Affordable Billion-scale Deep Learning Model Training
2022 IEEE/ACM Super Computing (SC) (CCF A)
ROSE: Cluster Resource Scheduling via Speculative Over-subscription
2018 IEEE International Conference on Distributed Computing Systems (ICDCS) (CCF B)
News & highlights
- Recipient of a HiPEAC Technology Transfer Award.
- Awarded a further £40,000 in non-dilutive seed funding through the AI SuperConnector programme.
- ACE3 selected to exhibit at AI UK 2025 in London.
- Awarded $25,000 in Google Cloud support as principal investigator for the AISC project.
- Co-awarded £16,000 from the International Strategy Fund for a Joint Research Centre, as co-investigator.
- Awarded £20,000 in non-dilutive seed funding through the first cohort of the AI SuperConnector programme, as principal investigator.
- Awarded £389,000 in funding, including £150,000 in EPSRC IAA support, to translate PhD research into the spin-out company ACE3 AI Ltd.