The rapid expansion and popularity of on-demand transportation services is transforming the urban mobility status-quo, shifting demand from personally owned and operated vehicles to the Mobility as a Service (MaaS) paradigm. One of the major limiting factors in this transformation has been the supply side complexities of operating a Mobility-on-Demand (MoD) service. With the potential for large fleets of autonomous vehicles to be available in the future, MoD providers will likely to be able to provide a more pervasive service at a cheaper cost, thereby attracting a larger percentage of the total transportation demand. While such a future is appealing in terms of more convenient, available and perhaps cheaper personal mobility, one needs to also consider the network level consequences of this transformation. In particular, there are many questions regarding the potential negative externalities arising from induced demand and the impact of empty vehicle rebalancing (or deadheading). While simply replacing private vehicle trips with autonomous taxi rides is unlikely to be a scalable and sustainable solution, autonomy does provide a number of opportunities for enabling high capacity MaaS solutions (e.g. on-demand micro-transit and transit feeder systems) that may not be economically feasible without vehicle automation. This talk will focus on some such opportunities, the algorithmic challenges in deploying the corresponding services, and ongoing work on addressing these algorithmic challenges.