I'm a physics doctoral researcher at the Institute for Quantum Gravity in the Friedrich-Alexander-Universität in Erlangen, Germany.

My main research (software related) interest is in developing and applying numerical methods which utilise deep learning techniques in both canonical and covariant loop quantum gravity.

Aside from that, I like fast code, so I also dabble around in code optimisation (particularly for massively parallel HPC codes). I am interested in GPU level performance engineering as well as accelerating and scaling CUDA applications to multiple GPUs/nodes, both Python and CUDA C/C++.

I am also interested in both node and core performance engineering, and I spend free time optimising low-level routines just to see how far the hardware will let me go. I enjoy understanding how the compiler thinks, (which means reading assembly when it doesn't) and finding a few extra nanoseconds in performance gains no one was asking for...

Whenever there is free time, I like to play around with, cryptography particularly post-quantum cryptography algorithms.

I am currently the lead developer of neuraLQX (to be released soon), which is a python package built on NetKet for the purpose of utilising neural network quantum states to solve constraints of canonical loop quantum gravity systems.