Galaxy imaging and redshift surveys, designed to measure gravitational lensing and galaxy clustering, remain the most powerful probes of large-scale structure. Such surveys constitute a significant fraction of current and next-generation projects in the cosmology community (e.g. DES, HSC, LSST, eBOSS, DESI, EUCLID, WFIRST). The statistical power of these experiments requires significantly improved understanding of astrophysical and observational effects. In this talk, I will focus on two important astrophysical processes which contribute systematic uncertainty but also contain a potential wealth of information. First, correlations in the intrinsic shapes and orientations of galaxies, termed “intrinsic alignments” (IA), are an important systematic in weak lensing. I will describe efforts to improve analytic modeling of IA as well as empirical work to constrain IA in lensing measurements. Second, redshift-space distortions (RSD) alter the observed shape of galaxy clustering and are a powerful probe of structure growth. I will describe how our sophisticated model, based on the matter phase-space distribution function, is able to effectively separate RSD from the geometric effects that probe the expansion of the universe and dark energy.