Baryon Acoustic Oscillations (BAO) datasets offer interesting possibilities for inferring cosmological parameters, due to their unique properties of being relics of the pre-recombination Universe. Since the BAO measurement are done by tracking the spatial distribution of millions of galaxies, the statistical effects and the assumptions on the model play significant role when the BAO data is used to constrain cosmological parameters. In this proceeding, we review our recent articles in which we tried to minimize some assumptions when we used BAO datasets to fit different cosmological models. In one of them, we use a BAO dataset along with cosmic chronometers, supernovae, gamma-ray bursts and quasars but we take the sound horizon as a free parameter. In the other, we take a BAO dataset along with a CMB distance prior and we take the combination $H_0 r_d$ as a free parameter. We discuss the results, and also the difficulties and the advantages of these approaches.