This project aims to mitigate model development and maintenance costs at all points along its lifecycle by improving modularity. We seek to improve efficiency using modular additions to the network and do so effectively in a manner that does not incur the cost of catastrophic interference. Such methods can, in time, be deployed to groups of modular adaptation across language, tasks and ground realities (factual updates). The ability to perform these in a composable way will enable important gains in the agility of model deployment in new domains and problem spaces.