Receptor-based detection of emerging environmental toxicants, including per- and polyfluoroalkyl substances (PFAS), is challenging due to their diverse chemical structures. We developed a novel method of dipping paper chips into water samples, using a set of biologically interacting molecules (BIMs) to evaluate the molecular interactions between PFAS and the BIM set. The System Automated Loading for Sample Analysis (SALSA) was introduced to ensure consistent and automated sample delivery onto paper chips. A smartphone captured video clips of capillary action, and cloud analysis evaluated the flow velocities, which were correlated with PFAS-BIM interactions. A comprehensive database of over 10,000 flow profiles was collected for each type of water sample, including mineral water, groundwater, and effluent wastewater. The optimized machine learning algorithm successfully classified PFAS presence (including 1000, 10, and 0.1 ppt) and its type. This integrated system demonstrates a reliable, portable, and scalable PFAS detection, and the observed PFAS-BIM interaction patterns suggest potential biological relevance.