We present multi-wavelength observations of two gap transients followed by the Carnegie Supernova Project-II and supplemented with data obtained by a number of different programs. Here in the first of two papers, we focus on the intermediate luminosity red transient (ILRT) designated SNhunt120, while in a companion paper we examine the luminous red novae AT 2014ej. Our data set for SNhunt120 consists of an early optical discovery, estimated to be within 3 days after outburst, the subsequent optical and near-infrared broadband followup extending over a $\sim$2 month period, two visual- and two near-infrared wavelength spectra, and Spitzer Space Telescope observations extending from early ($+$28 d) to late ($+$1155 d) phases. SNhunt120 resembles other ILRTs such as NGC 300-2008-OT and SN 2008S, and like these other ILRTs, SNhunt120 exhibits prevalent mid-infrared emission at both early and late phases. From the comparison of SNhunt120 and other ILRTs to electron-capture supernova simulations, we find that the current models underestimate the explosion kinetic energy and thereby produce synthetic light curves that over-estimate the luminosity. Finally, examination of pre-outburst Hubble Space Telescope images yields no progenitor detection.

We present optical and near-infrared broadband photometry and optical spectra of AT 2014ej from the the Carnegie Supernova Project-II. These observations are complemented with data from the CHilean Automatic Supernova sEarch, the Public ESO Spectroscopic Survey of Transient Objects, and from the Backyard Observatory Supernova Search. Observational signatures of AT 2014ej reveal that it is similar to other members of the gap-transient subclass known as luminous red novae (LRNe), including the ubiquitous double hump light curve and spectral properties akin to the LRN SN 2017jfs. A medium-dispersion, visual-wavelength spectrum of AT 2014ej taken the Magellan Clay telescope, exhibits a P Cygni H$\alpha$ feature characterized by a blue velocity at zero intensity of $\approx 110$ km s$^{-1}$ and a P Cygni minimum velocity of $\approx70$ km s$^{-1}$, and which we attribute to emission from a circumstellar wind. Inspection of pre-outbust Hubble Space Telescope images yields no conclusive progenitor detection. In comparison with a sample of LRNe from the literature, AT 2014ej lies among the brighter end of the luminosity distribution. Comparison of the ultra-violet, optical, infrared (UVOIR) light curves of well-observed LRNe to common-envelope evolution models from the literature, indicates the models under predict the luminosity of the comparison sample at all phases and also produce inconsistent time-scales of the secondary peak. Future efforts to model LRNe should expand upon the current parameter space explored and therefore may consider more massive systems and a wider range of dynamical timescales.