Fitting high-S/N Spectral Energy Distributions from cross a range of wavelengths allows us to better resolve degeneracies between quantities like the star formation rate and dust. This in turn allows us to robustly extract information about the different stellar populations that comprise a galaxy's Star Formation History (SFH). Using the Dense Basis method (Iyer & Gawiser 2017), we reconstruct the SFHs with uncertainties for a large sample of galaxies from the CANDELS catalog. Using Gaussian Process Regression, we encode the parameters describing these SFHs in a functionally independent form. This gives us more robust estimates for quantities like Stellar Masses and SFRs, that directly depend on the SFH. In addition, these SFHs can be used to answer questions like the time at which a galaxy's star formation peaked, and how many major episodes of star formation occurred in a galaxy's past, allowing us to go beyond the traditionally estimated 'Galaxy Age', which is often poorly constrained. They also allow us to probe the high-redshift low-stellar mass regime of the SFR-M* correlation by constructing trajectories in SFR-M* space for each galaxy.