An energy-efficient real-time processor that enhances R-waves in an electrocardiogram (ECG) signal is presented. The processor leverages a non-linear filter that models a system consisting of a particle inside a monostable well potential. The system is known to facilitate stochastic resonance (SR), where additive noise helps improving detectability of a weak signal. The processor is designed using analog signal processing techniques for simplicity of implementation and energy efficiency. Based on the schematic-level circuit simulations on the MIT-BIH arrhythmia database, the processor achieves an average sensitivity of 99.78% and an average positive predictivity of 99.65%. The power consumption excluding the bias circuitry and the thresholding stage is 3.75 nW. The results serve as a proof-of-concept demonstration towards facilitating SR in practical signal enhancement and detection scenarios with limited power budgets.