I'm a PhD student at the Jodrell Bank Centre for Astrophysics, University of Manchester, where I spend most of my time trying to convince contaminated radio data to tell me something true about the early universe. My work centres on the Epoch of Reionization and Cosmic Dawn, using Bayesian methods to pull real cosmological signal out from under instrumental systematics that would otherwise swamp it.
Along the way I've ended up thinking about uncertainty as much as I think about answers. How confident should we actually be in an estimate? How do you build that into a model rather than bolt it on afterwards? That's pulled me into machine learning and statistical modelling more broadly, including a stint as a data scientist at American Express, building models that fed real marketing decisions.
These days I am more interested in the intersection of AI/ML and Bayesian statistics, since I believe combining some of the best tools we have at our disposal can extract new information from old data.
If you'd like to chat, send me an email!

Research Interests

• Cosmology & Early Universe: Epoch of Reionization, Cosmic Dawn, Inflation
• Bayesian inference & uncertainty quantification
• Machine learning for scientific inference
• Cosmological simulations

Latest Work

• Invited talk on "Bayesian Pathways to Cosmic Dawn and Reionisation: From HERA Visibilities to Galaxy Cross-Correlations" at IIT Indore, India (July 2026)
• Sohini Dutta,Phil Bull,Jacob Burba, Michael J. Wilensky, Zheng Zhang, Ainulnabilah Nasirudin Bayesian power spectrum estimation with modelling of systematic effects in delay-fringe rate space
• From ANN to BNN: Uncertainty-aware emulators of 21-cm summaries — arXiv
• RHINO: A large horn antenna for detecting the 21 cm global signal — arXiv

See more on Publications & Codes or Conferences & Workshops.