About me

I work where state-of-the-art machine learning meets fundamental gravitational physics, using contrastive learning to turn gravitational-wave data into physical insight.

I am a postdoctoral researcher at ETH Zürich, working on contrastive learning for gravitational-wave science. I completed my PhD in 2025 at GRAPPA, University of Amsterdam.

My work spans the gravitational-wave spectrum and the communities built around it: ground-based detectors with LIGO–Virgo–KAGRA and the Einstein Telescope, space-based detection with LISA, pulsar timing arrays, and electromagnetic follow-up with GRANDMA.

I enjoy building and using contrastive learning algorithms. I have worked extensively on neural ratio estimation, in particular marginal and targeted inference, and I am now also developing contrastive learning for representations of gravitational-wave data.

I maintain these open-source projects:

• peregrine: neural ratio estimation for gravitational-wave inference.
• praxis: an AI harness that runs a large language model as a small research lab following the scientific method.

For a full list of the open-source software I have led or co-developed, see Software.

Before Amsterdam, I did my MSc at the Zentrum für Astronomie, Heidelberg University, and my BSc at the University of Delhi.

More detail is in the research and publications sections. For collaborations or questions, get in touch.

News