Ciarán Gilligan-Lee completed his Master’s degree at Perimeter Institute for Theoretical Physics. Now he’s improving automated decision-making as the head of Spotify’s Causal Inference Research Lab.
Perimeter Institute alumni have gone on to a wide variety of roles after leaving the Institute. After finishing his doctorate at the University of Oxford, Ciáran Gilligan-Lee’s path was heading toward academia, but instead he joined a health startup and then Spotify, interested in seeing if he could apply causal inference to real-world problems. We reached out to Ciarán to learn more about his journey.
This interview has been lightly edited for clarity and length.
What is your current role, and how are you trying to push boundaries in your field?
I am the head of Spotify's Causal Inference Research Lab. I lead an international team of scientists conducting research in causality and machine learning to make automated decision-making more trustworthy, efficient, and scalable.
Machine learning is a powerful tool capable of providing scalable solutions to complex problems. But it can only learn from correlation in data. This is a major drawback, as understanding cause and effect is crucial for decision-making – and correlation is not causation. My work addresses this challenge by creating new models that combine causal reasoning with the scalability of machine learning, yielding accurate decision-making at scale.
Spotify isn’t the only company racing to build models that can reason causally. Meta, Amazon, and Uber are developing the technology. However, my lab at Spotify was the first to create scalable causal machine learning models. My work isn’t just impactful for Spotify, it has been applied to many decision-making scenarios: “Should we protect wells in Kenya to improve health?” “What drug is best for this patient?” “I was denied a loan online, why?”
What brought you to where you are now?
From 2012 to 2013, I was a Perimeter Scholars International (PSI) student at Perimeter Institute, where I graduated as Valedictorian. It was there that I was first exposed to causal inference by Dr. Rob Spekkens while working on my PSI essay with him. Looking back, this was a transformative moment that has had a huge influence on my career. The work Rob and I started at Perimeter Institute was eventually published in the Journal of Causal Inference, and lit a fire in me to see how these techniques could be utilized to solve real-world problems.
After my DPhil at the University of Oxford, I received research funding from the EPSRC in the United Kingdom to utilize causal inference to develop quantum cryptographic protocols for large-scale, real-world quantum networks. Towards the end of this project I was offered a permanent Assistant Professor position at a leading United Kingdom university. At this crucial juncture I decided against staying in academia, and decided to try and see if I really could apply causal inference to solving real-world problems. I took a role in a tech healthcare startup, where I led a research lab in developing causal inference tools to address problems such as automated medical diagnosis. From there, I never looked back.
What are you passionate about?
The most important thing about me is that I am a dad. The thing I am most passionate about is showing the world to my two year-old son. When we have children, we must re-explain the world, and in doing so come to know the place for the first time. I am finding this a numinous experience.
In a distant second place is my love and passion for brewing (and drinking!) coffee. I seem to spend too much money exploring how different coffee bean varieties, when grown, processed, and roasted in different ways, yield different flavours. There is also a lot of physics in how coffee is brewed, and I am enjoying getting hands-on experience with the nuances of percolation.
Finally, I am also passionate about public communication of science and technology and how best to enhance the public's understanding of the latest developments in these fields. I have written multiple feature articles for New Scientist, the world’s most read weekly science magazine on topics from AI and causality to quantum computing.
How has your work impacted your industry and community?
The connection between cause and effect is how we first learn about the world. But, surprisingly, modern machine learning is blind to cause and effect, relying only on correlations extracted from data.
My research has focused on combining concepts from fundamental physics with deep learning to develop algorithms for causal reasoning. My main achievement has been to apply these algorithms in diverse areas: privacy, healthcare and decision-making. I’ve used my work to dramatically reduce the rate of automated medical misdiagnoses, develop quantum cryptography for large-scale quantum networks, and make automated decision-making more efficient and trustworthy.
How do you give back to your community?
For novel discoveries in AI to have impact, I believe academics must apply them to problems facing society. In 2018 I was offered a tenured position at a prestigious university, but chose to enter industry to apply AI to solve real-world problems. While working towards this goal, I co-supervise and mentor PhD students from University College London in applying advances in AI to physics.
I am also passionate about enhancing the public's understanding of the latest developments in AI, as well as building public trust in AI. I have written multiple feature articles for New Scientist. I am also regularly interviewed as an AI expert, and in 2025, I gave a TEDx talk that overviewed the potential of AI and what work is still needed to make it capable of solving important real-world problems in a trustworthy manner to a large audience in Dublin, Ireland.