10/2014-04/2018, PhD, supervised by Satish Singh and me. I had known Shihao since his M2 at IPGP. He joined the marine geosciences team and started developing localised time-lapse waveform inversion. We spent a whole year (re)formulating the representation theorem for a small perturbation, which could finally be interpreted as a perturbation theory. Even though I had experience with this kind of theory, it was with Shihao that I could theoretically connect many existing techniques in theoretical and numerical seismology (the appendix of our first paper is still one of the works I am most proud of). With Dmitry and Satish, we advanced the methodology, and I plan to use the code we put together to image magma chambers with Milena, who joined IPGP in September 2021. As Shihao has always been interested in rotational seismology, he did his postdoc with Heiner Igel in Munich, taking part in several projects around this new technology. He is now a professor (CPJ) at Université Gustave Eiffel.