Taavet Kalda

Taavet Kalda is a PhD student at the Max Planck Institute for Astronomy in Heidelberg, working on dynamical modelling of the solar neighbourhood of the Milky Way with machine-learning methods such as normalizing flows. He completed an Integrated Master's in Mathematical and Theoretical Physics at the University of Oxford in 2021, and earlier worked as an engineer at TalTech on an epidemiological COVID-19 simulation and as a data analyst at the Atlas Fellowship. He represented Estonia at the International Olympiad in Informatics (IOI 2017, silver medal) and is active in physics/astronomy olympiad outreach and problem-book compilations.

Notes7

• Quant employment verification: UNRESOLVED. The only evidence of quantitative-finance employment is a single index-only (LinkedIn self-report) entry, 'Quantitative Research Intern at Jump Trading Group, 2020', with no month precision and no independent primary-source corroboration found; Jump Trading is a genuine proprietary trading / quantitative market-making firm, so the role title and employer are plausible, but CONFIRMED status requires an actual internship plus a supported identity evidenced beyond an unverified index entry.
• Index-only (LinkedIn self-report): Quantitative Research Intern at Jump Trading Group, dated '2020' with no month precision; no non-index primary source yet located.
• Doctoral Researcher at the Max Planck Institute for Astronomy (mpia.de), Heidelberg, in the Galaxies and Cosmology department, supervised by Gregory M. Green; research on dynamical modelling of the solar neighbourhood with machine learning/normalizing flows, stated on track to graduate May 2026.
• Worked as a Data Analyst at Atlas Fellowship (2022-2023) and as an Engineer at TalTech developing a full-scale dynamical epidemiological COVID-19 simulation for Estonia (2021-2022); also tutored the Saudi Arabian national physics team via Mawhiba (2018) - none of which are quantitative-finance roles.
• LinkedIn lists an Integrated Master's in Mathematical and Theoretical Physics at the University of Oxford, 2017-2021.
• Represented Estonia at IOI 2017, earning a silver medal; Codeforces handle 'poomas'.
• First author on 'Deep Potential: Recovering the gravitational potential from a snapshot of stellar coordinates' (arXiv 2507.03742), and co-author on 'Deep Potential applied to a simulated barred galaxy' (MNRAS 527/4, 2023); also a co-author of an AAPT American Journal of Physics paper on the first Global e-Competition on Astronomy and Astrophysics.
• Presented a seminar 'Deep Potential: Determining the Milky Way's gravitational potential without selection functions' at Westlake University on 2025-12-22.
• Won the Estonia-Finland physics competition while a school student.
• Featured among Estonian young scientists abroad in a January 2025 Research in Estonia profile of the Estonian Academy of Sciences 'Estonian Young Scientists Abroad' conference.
• Maintains a personal academic site at tkalda.ee / majakas.github.io and a public GitHub account (github.com/Majakas).