We report on NICER X-ray monitoring of the magnetar SGR 1830−0645 covering 223 days following its 2020 October outburst, as well as Chandra and radio observations. We present the most accurate spin ephemerides of the source so far: ν = 0.096008680(2) Hz, Hz s−1, and significant second and third frequency derivative terms indicative of nonnegligible timing noise. The phase-averaged 0.8–7 keV spectrum is well fit with a double-blackbody (BB) model throughout the campaign. The BB temperatures remain constant at 0.46 and 1.2 keV. The areas and flux of each component decreased by a factor of 6, initially through a steep decay trend lasting about 46 days, followed by a shallow long-term one. The pulse shape in the same energy range is initially complex, exhibiting three distinct peaks, yet with clear continuous evolution throughout the outburst toward a simpler, single-pulse shape. The rms pulsed fraction is high and increases from about 40% to 50%. We find no dependence of pulse shape or fraction on energy. These results suggest that multiple hot spots, possibly possessing temperature gradients, emerged at outburst onset and shrank as the outburst decayed. We detect 84 faint bursts with NICER, having a strong preference for occurring close to the surface emission pulse maximum—the first time this phenomenon is detected in such a large burst sample. This likely implies a very low altitude for the burst emission region and a triggering mechanism connected to the surface active zone. Finally, our radio observations at several epochs and multiple frequencies reveal no evidence of pulsed or burst-like radio emission.