Avish Vijayaraghavan
About me
I’m a scientist-engineer interested in AI and pharma, based in West London. My mission is to build out a pharma model that works with the NHS. My research interests cover interpretability (models that we can understand), multi-modal learning (models that combine multiple data modalities together), and deploying models in real clinical settings.
I’m currently doing a PhD in AI for Healthcare at Imperial working on clinical multi-omics models for IPF, a rare lung disease. Before that, I did a master’s in precision medicine at UCL, and a bachelor’s in maths and computer science, also at Imperial.
On the careers side, I’ve done a few internships. My first was with Microsoft Research’s BioML team, working on interpretable neural networks for Parkinson’s proteomics - this collaboration is ongoing. My second was between the data science teams of the NHS and Great Ormond Street Hospital, building annotation workflows for paediatric kidney reports using LLMs. I recently did a third, again at Great Ormond Street Hospital, this time in their partnership with Roche, using LLMs to analyse paediatric cancer reports.
For my aims of aligning pharma and the NHS, I need an understanding of the broader ecosystem of healthcare and pharma: finance, economics, policy, and education. To this end, I’m a fellow at No Patient Left Behind, an organisation dedicated to making pharma innovation affordable via health insurance reform in the US. My notes from the course readings are here. I’ve also tutored GCSE, A-Level, and undergraduate students, and done consulting work for life sciences and tech companies.
My CV is available here. If you’d like to get in contact, drop me a message on LinkedIn, X, or email. Outside of work, I occasionally write, make videos, and make tunes.