I obtained my PhD in Computer Science at the University of Cambridge in 2019, studying algebraic routing theory. Since then I have worked at both Curtin University, Western Australia, and at Heriot-Watt University, Scotland on a variety of projects. In my spare time, I'm the chief maintainer of the standard library for the Agda theorem prover prover, and occasionally like to dabble in computational problems in geology.

In general I'm interested in formal proofs of correctness of software. While in the past I have looked at the theory behind distributed algorithms and network routing protocols, my current focus is on the correctness of neural-symbolic programs - systems with both classical software and machine learning components. To that end, with a variety of collaborators, I am currently developing Vehicle, a tool for the end-to-end verification of such systems. Using Vehicle, you can train your network to satisfy logical specifications, rigourously check that the specification holds and then export the proof to an interactive theorem prover.

I'm always open to collaborations and applications so please do get in touch if you have neural networks whose correctness is important to you, or if you're interested in persuing a PhD in this area!