Using the NuMI beam at Fermilab and the NOvA near detector, we study the process by which a muon neutrino elastically scatters off an electron in the detector to produce a very forward going electromagnetic shower. By comparing dE/dx for various particle hypotheses for both longitudinal and transverse directions in a multilayer perceptron neural network, we trained a Particle ID algorithm to identify the scattered electron in an inclusive dataset. Muon-neutrino-on-e elastic scattering provides a clean, purely leptonic process free from nuclear effects for understanding neutral current scattering and constraining the NuMI beam flux. Also, this technique can be applied in two broad areas of beyond the standard model physics: a large neutrino transition magnetic moment and light dark matter particles produced in the NuMI target, both of which would create an energy dependent enhancement in the elastic scattering cross section.