The project
ArqueoGal, in full "Unraveling the evolutionary history of the Milky Way with Galactic Archaeology", is an FCT exploratory research project led by Tiago L. Campante at IA, with the Universities of Padova and Bologna as partners, running from February 2026 to August 2027. It sets out to build a high-resolution map of the Galactic disc in age, chemistry and kinematics from the red giants whose oscillations TESS has measured: asteroseismic ages from TESS, astrometry and orbits from Gaia, and abundances from APOGEE, with a machine-learning tool to classify the stellar populations.
My part
I am a co-investigator (team member) on the project and I develop its machine-learning pipelines, with allocations on the stellar population classification and Galactic modelling tasks. That work has three repositories. ArqueoGal itself is the platform: it prepares and corrects the observational catalogues, computes the kinematics and assembles the project's deliverables under an explicit release policy, with provenance on every file.
It contains no model. JANUS, in its own repository, owns the Gaia XP input chain, the normalising-flow ensemble and the per-star reliability. Starfold, open source, owns the population search and its recurrence evidence. ArqueoGal hands each tool its input and receives a scored table back, so the three can be tested and replaced on their own.
Where it runs
The two FCT advanced-computing projects on Deucalion and MareNostrum 5, which I hold as principal investigator, run the JANUS side of this work. My PhD fellowship since September 2026 is paid by the project.