Ultra-light axions of mass around 10^{-22} eV are a promising dark matter candidate well motivated by high energy physics. In order to constrain these dark matter candidates, we need our full arsenal of observations. I will discuss constraints on the axion mass coming from small-scale CMB measurements from telescopes like the Planck satellite, the clustering of galaxies from the Baryon Oscillation Spectroscopic Survey (BOSS) and weak gravitational lensing constraints from the Dark Energy Survey (DES). These complementary approaches help disentangle different physical affects of axions, but require nonlinear modelling of structure formation. We model the effect of FDM on structure formation by implementing effects on the halo mass function and concentration parameter found in N-body simulations into the halo model of non-linear structure formation. I will also discuss the future of axion research in the observational context.