I initially contributed as a software engineer, writing scalable data pipelines for processing Protein Data Bank entries. Later I become a "biology advisor" to the team, helping to assess the biological relevance of our predictions and handling collaboration requests from external scientists. I was particularly involved in the process of making AlphaFold publicly available. During this period I wrote a paper on the model's output applied to the whole human proteome, and worked with EMBL-EBI to produce a database of structure predictions covering human plus key model organisms.

My work now typically involves model evaluation - assembling appropriate test sets in order to judge whether a model is making progress on some biological problem. I still love discussing possible collaboration opportunities with the wider biology community!