To meet evacuation needs from carless populations who may require personalized assistance to evacuate safely, we propose a ridesharing-based evacuation program that recruits volunteer drivers before a disaster strikes, and then matches volunteers with evacuees who need assistance once demand is realized. We optimize resource planning and evacuation operations under uncertain spatiotemporal demand, and construct a two-stage stochastic mixed-integer program, first with an expectation-based objective and then a probabilistic constraint, to ensure high demand fulfillment rates. In addition to optimizing the sample average approximations of the two formulations, we study a heuristic approach that is able to provide quick, dynamic and conservative solutions. We demonstrate the performance of our approaches using five test networks of varying sizes based on regions of Charleston County, SC, an area that experienced a mandatory evacuation order during Hurricane Florence. We utilize demographic data and hourly traffic count data from the day of evacuation to estimate the demand distribution. The risk-neutral formulation ensures an average 97\% demand fulfillment rate for all test cases and complete fulfillment rates for larger instances at peak-demand time. For smaller instances at off peak time, the risk-averse approach achieves better reliability results.