This paper introduces a fast artifact recovery algorithm (FARA) that uses electrochemical impedance spectroscopy to model the electrode-tissue interface and design an optimum stimulation waveform to minimize the residual artifact duration in single-cell resolution neural interfaces. Results in saline solution with a custom PCB and a 30μm diameter microelectrode array show a worst-case artifact recovery time of 160μswhen measured from the end of the working phase (anodic 500 nA, 250μs). On average, the proposed algorithm provides an 81% improvement over a triphasic charge-balanced stimulation waveform.