Hi! I'm Ollie, a SABS PhD student funded by the EPSRC. Previously, I completed my MBiochem here at Oxford, and now work on applying ML to therapeutic antibody discovery.

Deep learning for therapeutic antibody optimisation

I'm a first year DPhil student, co-supervised by Yee Whye Teh and Jakob Foerster. Before Oxford, I was a Software Engineer at Meta for over 3 years. I completed my Bachelor and Masters in Computing at Imperial College London.

I am interested in improving the generalizability and robustness of ML models. Particularly, I'm interested in structural causal models, latent variable models and causal structure learning.

I am particularly interested in causal inference, uncertainty quantification, and enhancing the robustness of machine learning methods, especially for policy learning, evaluation, and healthcare applications. Before joining the StatML CDT, I worked as an AI Applied Scientist at Microsoft New England Research Center, where I gained two years of experience addressing real-world challenges. I hold a master's degree in Computational and Mathematical Engineering from Stanford University. My passion for causal inference drives my commitment to developing statistical and machine learning solutions that are both rigorous and applicable in critical domains like healthcare.

• Causal Inference
• Robust Statistics
• Uncertainty Quantification

I am a second year DPhil student under the supervision of Professor Christina Goldschmidt. I previously completed my BSc in Honours Mathematics, and MSc in Mathematics and Statistics, at McGill University.

Topics in random graphs and trees, mixing times of Markov chains, and interacting particle systems.

I am half English (Wessex) and half Italian (Sicily) and to make things more confusing I have lived in a few other places: New Hampshire, New Zealand (Dunedin, close to Rohan and Lothlórien) and Copenhagen. I collect hobbies of the nerdier kind: everything for me seems to turn into an intriguing experiment and in my free time, I sometimes code for fun, rewire the electrics in the house, 3D print something utterly useless and cursed or solder components for an equally useless and janky Raspberry-Pi–based sensor.

I am a computational biochemist working in drug design.

My loyalties lie with molecular thermodynamics, but I am starting to fall the allure of deep learning.

I am currently working on Fragmenstein, github.com/matteoferla/Fragmenstein, a tool that stitches molecules together and reanimates them in order to help assess follow up compounds in fragment-based drug discovery.

Previously, I have developed Michalenglo (michelanglo.sgc.ox.ac.uk/), a tool allowing the sharing of representations of 3D structures of protein with interactive annotations, and Venus (venus.cmd.ox.ac.uk/venus), a tool that combines different sources of information in order to determine what could be the effect of a missense variant beyond simple destabilisation.

For other projects of mine, visit my github: github.com/matteoferla

Additionally, I have a science blog (blog.matteoferla.com) ranging from technical tutorials to monographic discussions/rants.

Jeffrey is from Hong Kong and studied at Pui Ching. He spent his undergraduate years at the University of Waterloo in Canada, where he majored in pure mathematics, statistics and combinatorics & optimisation. Jeffrey also studied at the University of Warwick for a year. Before his DPhil, he obtained his master’s degree in mathematical sciences (OMMS) from Balliol College, Oxford. As a Christian, he follows Christ the Lord as his light and his salvation. In his spare time, Jeffrey enjoys reading novels and collecting small trinkets of sentimental value.

• Causal inference
• Instrumental variables estimation
• Mendelian randomisation