Hi there, I’m Christina 👋!
I’m a research engineer based in London, with a particular interest in training machine learning research algorithms at large scale, and applying them to impactful real-world applications.
Most recently, I contributed to building DeepPack as a research engineer at InstaDeep. This involved applying reinforcement learning to BinPacking – an NP-hard combinatorial optimization problem – in a way that is computationally efficient and generalisable to unseen client data.
Prior to this, I was a research fellow in the Empirical Inference group at Max Planck Institute for Intelligent Systems, where I worked on distributionally robust optimization with Dr. Jia-Jie Zhu and Prof. Bernhard Schölkopf. This was published as “Adversarially Robust Kernel Smoothing” (oral presentation at AISTATS 2022).
I completed my MSc in Computational Statistics and Machine Learning at University College London, fully supported by a Cisco AI scholarship. As part of my MSc thesis, I investigated how the syntactic structure of language can be utilised to improve sample efficiency and generalisation in instruction following, under the guidance of Dr. Tim Rocktäschel and Minqi Jiang (UCL DARK lab).
Before transitioning to AI, I was part of an analytics start-up incubated by fashion e-commerce firm Farfetch. I was involved in all aspects of building a Market Intelligence software and launching Farfetch’s first loyalty scheme.
For my first degree, I studied Civil, Energy, and Environmental Engineering at the University of Cambridge, where I received a BA and MEng with Distinction. For my Master’s thesis, I performed research on design parameters for automated bridge inspection using Augmented Reality, supervised by Dr. Brilakis in the Construction Information Technology research lab.
My personal interests include classical piano, cycling, and photography – I particularly enjoy capturing foreign cultures, mountains, and our galaxy.