Time-domain astrophysics is entering a golden age with discovery rates increasing exponentially thanks to combining public and private surveys such as the Young Supernova Experiment and the Zwicky Transient Facility. However, the number of events isn't the key metric we should optimize for, but rather the number of discoveries that maximize science return, and here there remains much to do. We still do not understand the progenitor systems or explosion mechanisms of most classes of supernovae, let alone how these events feedback into the ISM and impact their host-galaxies. I will discuss how we've been leveraging data-intensive science and artificial intelligence to study the host environments of supernovae, identify the most common and the most anomalous objects, follow them up spectroscopically, and use this information to improve the utility of type Ia supernovae as cosmological probes of dark energy. I will also highlight how we are taking the lessons learned from combining the heterogenous data from the current generation of surveys, and applying them to the nascent field of multi-messenger astrophysics to enable time-domain astrophysics with a latency of minutes, instead of days. Finally, I will outline how I believe the University of Illinois can play a key role as a hub for the emerging time-domain ecosystem with the Vera C. Rubin Observatory, and the Nancy Grace Roman Space Telescope.
