During an extended internship at Amazon AI, I worked on the theory of deep learning. We designed new neural net optimisers that get the best of both gradient adaptivity and compression, along with informative theoretical guarantees.


So, you’re working on an important problem that might one day really help people. Why didn't you tell the world? I helped start Caltech Letters to allow the Caltech community to do exactly that.


For my Masters' project I worked with Richard Turner and Máté Lengyel in the Engineering Department at Cambridge. We worked on the neural sampling hypothesis. The project looked at how the brain might use sampling to perform Bayesian inference.


I worked for British startup Improbable who are abstracting away the problem of building giant simulations. I was a software engineer on the Living Cities project. We built a detailed simulation of traffic flow in my hometown of Manchester.
During a wonderful Angeleno summer in Caltech's famous SURF program, I worked on ways to speed up Markov chain algorithms by running them on a quantum computer. Read my SURF report here.


I was lucky to be president of the Trinity College Students' Union. I focussed on outreach initiatives in Trinity, resurrecting the TCSU's web presence and pushing the college to take outreach more seriously. Read the report I wrote for Trinity's Annual Record.


In high school I was fascinated by emergence and the distributed intelligence of swarms. During one Mancunian summer I created Project Swarm. I drew inspiration from a brief stint in Steve Furber's lab at Manchester University, where researchers were building a computer with spiking neural network architecture.