Kieran is a PhD student in audio representation learning applied to computational ecology. In particular examining the utility of generative AI for interpreting ecoacoustic inference. By leveraging prior knowledge about the domain as inductive bias, he is developing methods to interpret downstream model predictions, disentangle factors of variation that are not ecologically relevant and quantify uncertainty in predictions. His work is a blend of machine listening, signal processing, mathematics, music and ecology. Kieran’s current research is funded by the Leverhulme Be.AI scholarship programme at the University of Sussex.
Before his PhD, Kieran honed his skills as an open source software engineer and member of several horizontal technology co-operatives. Largely working with distributed data structures and single board computers, he focused on building privacy-by-design and human-centric technologies. During this period Kieran sourced grant funding from many prominent sources including the EU’s Horizon 2020 research and innovation programme, the Ethereum Foundation, the Prototype Fund, and the Open Technology Fund.