The scientific computing landscape has evolved dramatically in the past few years, with new schemes for organizing and storing data that reflect the growth in size and complexity of astronomical data sets. In response to this changing landscape, we are, over the next two years, deploying the next generation of the Montage toolkit ([ascl:1010.036]). The first release (October 2015) supports multi-dimensional data sets (data cubes), and insertion of XMP/AVM tags that allows images to drop-in to the WWT. The same release offers a beta-version of web-based interactive visualization of images; this includes wrappers for visualization in Python. Subsequent releases will support HEALPix (now standard in cosmic background experiments); incorporation of Montage into package managers (which enable automated management of software builds), and support for a library that will enable Montage to be called directly from Python. This next generation toolkit will inherit the architectural benefits of the current engine - component based tools, ANSI-C portability across Unix platforms and scalability for distributed processing. With the expanded functionality under development, Montage can be viewed not simply as a mosaic engine, but as a scalable, portable toolkit for managing, organizing and processing images.