/ AbstractWe present optical photometric and spectroscopic observations of the Type~IIb supernova (SN)~2017ati. It reached the maximum light at about 27~d after the explosion and the light curve shows a broad, luminous peak with an absolute $r$-band magnitude of $M_{r} = -18.48 \pm 0.16$~mag. At about 50~d after maximum light, SN~2017ati exhibits a decline rate close to that expected from the $^{56}$Co $\rightarrow$ $^{56}$Fe radioactive decay, at 0.98 mag per 100 days, as usually observed in SNe IIb. However, it remains systematically brighter at late times by about 1--2~mag, exceeding the usual upper luminosity range of this class. As a result, modelling the light curve of SN~2017ati with a standard $^{56}$Ni decay scenario requires a large nickel mass of up to $\sim0.37\,M_{\odot}$ and still fails to reproduce the early-time light curve adequately. In contrast, incorporating additional energy input from a magnetar yields a significantly improved fit to the light curve of SN~2017ati, which would reduce the nickel mass to $\sim0.21\,M_{\odot}$, still close to the upper end of the range typically inferred for SNe~IIb. Comparing the fitted results of SN~2017ati with the known sample of SNe~IIb indicates that its luminosity evolution is best explained by a combination of neutron star spin-down energy and radioactive nickel deposition. From late-time nebular spectra of SN~2017ati, the luminosity of the [\Oi]~$λ\lambda6300,6364$ doublet implies an oxygen mass of $\sim1.82-3.34\,M_{\odot}$, and the combination of a [\Caii]/[\Oi] flux ratio of $\sim0.5$ with nebular spectral model comparisons favours a progenitor zero-age main-sequence mass of $\geq17\,M_{\odot}$.