We review recent advances and ongoing work in evolving the NASA/IPAC Extragalactic Database (NED) beyond an object reference database into a data mining discovery engine. Updates to the infrastructure and data integration techniques are enabling more than a 10-fold expansion; NED will soon contain over a billion objects with their fundamental attributes fused across the spectrum via cross-identifications among the largest sky surveys (e.g., GALEX, SDSS, 2MASS, AllWISE, EMU), and over 100,000 smaller but scientifically important catalogs and journal articles. The recent discovery of super-luminous spiral galaxies exemplifies the opportunities for data mining and science discovery directly from NED's rich data synthesis. Enhancements to the user interface, including new APIs, VO protocols, and queries involving derived physical quantities, are opening new pathways for panchromatic studies of large galaxy samples. Examples are shown of graphics characterizing the content of NED, as well as initial steps in exploring the database via interactive statistical visualizations.