Reconstructing the past of the Milky Way depends on the study of its metal-poor stars, which either have been formed in the Galaxy itself in the first billion years, or have been accreted through mergers of satellite galaxies over time. These stars are usually found in what is known as the Milky Way halo, a light — in terms of total mass —  stellar component which is usually made of stars whose kinematics significantly deviates from that of the Galactic disc.
In this talk, I will discuss how it has been possible to use the astrometric and spectroscopic data delivered by Gaia and complementary surveys  to shed light on the past of our Galaxy, through the study of its halo. Besides the discovery of the possible last significant merger experienced by the Milky Way, the use of 6D phase space information and chemical abundances allowed to reconstruct the impact this merger had on the early Milky Way disc, and the time it occurred, as well as to discover that some of the most metal-poor stars in the Galaxy possibly formed in a disc.  This last finding would imply that the dissipative collapse that led to the formation of the old Galactic disc must have been extremely fast.