Heya! I’m finishing my PhD in Engineering at the University of Oxford. My research focuses on developing explainable AI for time-series data in healthcare to predict heart attacks. I also develop AI to predict the spread of infectious diseases globally (like COVID-19), think about the interplay of the law and the need for explainability, and work on using AI safely to improve people's lives and health.

I completed my BSc in Electrical Engineering at NYU Abu Dhabi summa cum laude and worked on cool projects in Oman, India, the UK, and the United States.

After winning a Rhodes Scholarship as the first Bosnian in history, I moved to Oxford to pursue my MSc in Computer Science and my PhD. My masters thesis was awarded a distinction and it proposed a novel deep learning solution to improving CRISPR-guided gene engineering.

I'm passionate about the transformative power of technology and have worked in the disability space in the UAE and the US, have been on the Working Group for Trustworthy AI at Mozilla Foundation, was awarded a competitive Alan Turing Institute PhD Award to pursue independent research, and have co-founded a non-profit Oxford Diplomatic Society to mediate between strategic diplomacy and technology spaces.

These questions are currently on my mind (pls reach out if they resonate and you want to chat!):

• what makes an AI model explainable? what does 'explainable' even mean? to whom? can we verify that?
• how do we safely develop AI to predict tricky diseases?
• what are the new representation learning approaches for longitudinal health data?

research

My research draws on statistical and learning theory, time-series representation learning, graph modelling, clinical informatics, big health data science, and philosophy and law. I work directly with clinical collaborators globally to build AI systems that try to help improve or save lives. I also like to think about the bigger picture of how we integrate these models, what it means to 'explain' an AI decision, and how we make them more trustworthy in clinical practice. My publications can be found here.

news

* [Nov 2024] Published DySurv in JAMIA, proposing a novel dynamic risk profile prediction model for ICU survival

* [Oct 2024] Completed my first half-marathon at Oxford Half in 1h40m

* [Aug 2024] Presented DynaGraph, a contrastive and interpretable dynamic graph model for time-series, at 3 ICML workshops in Vienna

* [Jul 2024] Completed my strategy consulting placement through the Oxford Strategy Challenge with Criterium Solutions consulting on new mobile app development in South African market

* [Jul 2024] Published paper on ML and COVID-19 in Nature Scientific Reports, project supported by the Bill and Melinda Gates Foundation

* [Feb 2024] First anniversary as Junior Dean at Jesus College, Oxford, a role that continues to help me grow

* [Jan 2024] Joined the Turing Institute as an Enrichment PhD student to work on dynamic graph modelling and expand my network with leading AI researchers

* [Dec 2023] Presented DySurv, a CVAE model for survival analysis, at NeurIPS ML4H

contact