In this work, I describe the early attempts to engineering biosensors using a flavin-based fluorescent protein LOV. LOV proteins outperform fluorescent probes based on the green fluorescent protein (GFP) in anaerobic conditions and have the potential to serve as ligand- responsive allosteric reporters. Applications of LOV proteins to this end, however, have been limited due to a lack of understanding of engineerable sites within their protein structures. I systematically addressed this problem with the exhaustive library construction of domain- insertion chimeras using estrogen receptor ligand-binding domain (ERLBD) with iLOV. First, I extensively insert ERLBD into every single site of iLOV; the G480 site, which resides in the Hβ-Iβ loop within iLOV, retains 10% of the native iLOV fluorescence and exhibit a moderate increase in fluorescence upon ligand-binding (ΔF/F = 0.2). I also performed independent theoretical prediction of engineerable sites based on a homology-based model, an ensemble allostery-based model, and a machine learning model for protein function classification; all of which reached consensus on the experimentally validated sites. Next, based on the identified engineerable sites within iLOV, I swapped the ERLBD with an ATP-binding protein to engineer an ATP sensor. However, direct insertion didn’t yield a workable sensor; therefore, I introduced random short peptide linkers flanking the inserted domain. Furthermore, I developed a potassium cyanide (KCN)-based assay for ATP sensor screening. Although my screening assay didn’t allow the identification of a bona fide ATP sensor, I pointed out potential solutions regarding the control of cofactor availability, reduction of background fluorescence, and the use of an internal standard to achieve a streamlined screening process. The LOV-based biosensor is still in its early stage of development. My work presents a forefront framework to engineer LOV-based protein switches that may complement to GFP- based imaging in underexplored anaerobic biological systems.