Theoretical models of galaxy-AGN co-evolution ascribe an important role for the feedback process to a short, luminous, obscured, and dust-enshrouded phase during which the accretion rate of the SMBH is expected to be at its maximum and the associated AGN-driven winds are also predicted to be maximally developed. To test this scenario, we have isolated a text-book candidate from the eROSITA Final Equatorial-Depth Survey (eFEDS) obtained within the Performance and Verification program of the eROSITA telescope on board Spectrum Roentgen Gamma. From an initial catalog of 246 hard X-ray selected sources matched with the photometric and spectroscopic information available within the eROSITA and Hyper Suprime-Cam consortia, three candidates Quasars in the feedback phase have been isolated applying the diagnostic proposed in Brusa et al. (2015). Only one source (eFEDSU J091157.5+014327) has a spectrum already available (from SDSS-DR16, z=0.603) and it unambiguously shows the presence of a broad component (FWHM~1650 km/s) in the [OIII]5007 line. The associated observed L_[OIII] is ~2.6x10^{42} erg/s, one to two orders of magnitude larger than that observed in local Seyferts and comparable to those observed in a sample of z~0.5 Type 1 Quasars. From the multiwavelength data available we derive an Eddington Ratio (L_bol/L_Edd) of ~0.25, and a bolometric correction in the hard X-ray of k_bol~10, lower than those observed for objects at similar bolometric luminosity. The presence of an outflow, the high X-ray luminosity and moderate X-ray obscuration (L_X~10^44.8 erg/s, N_H~2.7x10^22 cm^-2) and the red optical color, all match the prediction of quasars in the feedback phase from merger driven models. Forecasting to the full eROSITA all-sky survey with its spectroscopic follow-up, we predict that by the end of 2024 we will have a sample of few hundreds such objects at z=0.5-2.