The hierarchical assembly history of our Milky Way galaxy has now been reconstructed in considerable detail, thanks to a combination of kinematic
and chemical evidence from Gaia and other surveys. As impressive as these results are, the Milky Way remains just one galaxy - how can we generalise
these results to other galaxies? How typical is the Milky Way? Even in the era of 30-40 m class telescopes, detailed studies of individual stars in galaxies beyond the Local Group will remain challenging, but I will argue that stellar clusters can play an important role as tracers of the early assembly histories of external galaxies. Interesting results have already emerged over the past decade from studies of Local Group GCs, including evidence of accretion in the M33 GC system and the discovery of the extremely metal-poor cluster EXT8 in M31. Future spectroscopic surveys with instruments such as the MOSAIC spectrograph on the ELT, following up on imaging by Euclid, Roman, and others, promise to enable such work on a much grander scale.

The colour–magnitude diagrams (CMDs) of star clusters underpin stellar astrophysics. They calibrate stellar evolution models, enable age-dating of stellar populations, and provide templates for interpreting unresolved galaxies. Yet the CMDs of young massive clusters show surprising features, such as split main sequences and extended turn-offs. This raises the question of whether we truly understand the mechanisms governing stellar distributions across the CMD.
In this talk, I will show how binary interactions and stellar rotation reshape CMD morphologies. Using observations of young massive clusters in the Magellanic Clouds obtained with HST and MUSE, we measure the spin velocities of large stellar samples and characterise the clusters’ binary populations. Our results reveal distinct populations of slow and fast rotators occupying different loci in the CMD. The detection of post-interaction binaries, such as blue stragglers and stripped stars, further enables us to constrain population synthesis models and to assess the contribution of massive star clusters to the cosmic rate of compact binary mergers observed via gravitational waves.

The cosmic web, consisting of large empty voids, cosmic walls, filaments and clusters, is characterised by multi-stream regions. A finite set of caustics can classify the geometry and formation histories of these regions. In particular, the cosmic walls correspond to the cusp caustic while the filaments correspond to either the swallowtail or the elliptic/hyperbolic umbilic caustics. In this talk, I will discuss how these structures are realised in our Universe with constrained N-body simulations.

Furthermore, I will discuss the first application of caustic skeleton theory to a cosmological redshift survey using the Manticore-Local simulations. With these Bayesian-constrained reconstructions of the Local Universe from the 2M++ galaxy catalogue, we extract the three-dimensional multi-scale caustic skeleton of two canonical web-like structures in our Local Universe, namely the Coma Cluster and the Pisces-Perseus ridge, as they represent the most prominent cluster node and filamentary artery in the nearby Universe. The Caustic Skeleton network of caustic singularities accurately reproduces the observed large-scale organisation of galaxies in redshift space for one of the Manticore realisations. Remarkably, while the Coma cluster seems to be governed by swallotail filaments, Pisces-Perseus is dominated by umbilic filaments.

The intracluster light (ICL) is a uniquely luminous probe of galaxy cluster halo shape and orientation, but it's extremely low surface brightness has historically made robust morphological measurements challenging. In this talk, I present ICL shape measurements from Euclid's Quick Release (Q1) for nearly 200 clusters at 0.1 ≤ z ≤ 0.8, selected from the eROSITA All-Sky Survey and the Dark Energy Survey. The ICL ellipticity distribution peaks at e ≃ 0.5, in close agreement with halo ellipticities from strong and weak lensing — supporting the view that the ICL traces the cluster's large-scale structure. Radially, BCG-dominated cores are relatively round, while the diffuse component becomes progressively more elongated before plateauing at large radii. By stacking clusters in redshift bins, we extend these measurements to larger cluster-centric radii and find no detectable evolution with redshift — a result that challenges theoretical expectations. Comparing to the Hydrangea simulations, we show that the ICL's measured shape depends sensitively on how it is defined and extracted: only fully forward-modelled mock observations reproduce the observed distribution, underscoring that realistic mock observations are essential for direct comparison with theory. This methodology provides a benchmark for testing hydrodynamical simulations, and forthcoming Euclid data releases will enable substantially tighter constraints.