Author an optimization model — including a model under uncertainty, with random variables, expectations, conditional value at risk and chance constraints — in a native R interface, and solve it with the 'Quicopt' service < https://quicopt.com>. Models are encoded to the service's wire format and solved remotely; no solver is installed locally. Random variables carry distributions or empirical scenario columns taken directly from a data frame, so observed history becomes a stochastic program in one call.