Lev Tankelevitch

À propos

I’m a behavioural scientist exploring ways to augment human agency in collaborative work and productivity. My current work uses metacognition—thinking about thinking—as a lens to understand and improve human-AI interaction, and to design Generative AI systems that help people think clearly and act more intentionally in collaborative work. My research approach is mixed-methods and reflects the intersection of behavioural science, human-computer interaction, and data science. I work in the Blended Reality for Effective Workflows (BREW)(opens in new tab) team, within the Collaborative Intelligence(opens in new tab) theme at Microsoft Research Cambridge (UK). I’m currently collaborating with Dr. Clara Colombatto(opens in new tab) as part of the AI, Cognition, and the Economy (AICE)(opens in new tab) pilot project on metacognition and trust in human-AI collaboration.

Before coming to Microsoft, I worked as a Research Advisor at the Behavioural Insights Team(opens in new tab), where I specialised in the design, implementation, and analysis of rapid online experiments, field experiments, and quasi-experiments across health, social care, and education. For example, I worked on a successful national-scale field RCT(opens in new tab) testing invitation messages to increase COVID-19 vaccinations in a cohort of ~2.3 million people.

I have an academic background in cognitive psychology and neuroscience. During my PhD at the University of Oxford, I researched the interaction between reinforcement learning and attention using behavioural experiments, brain imaging, and modelling.

To see all my past work, including in neuroscience, policy research, and science communication, see my personal website(opens in new tab).