<p>Single-molecule surface-enhanced Raman spectroscopy (SERS) has emerged as a transformative technique for detecting molecular vibrations with unprecedented sensitivity, especially for biomolecular analysis. In this study, we present a novel integration of AI-driven hyperspectral imaging and gold nanostars (AuNS) to enhance SERS analysis of amino acids at the single-molecule level. Gold nanostars were employed as plasmonic substrates to amplify Raman signals due to their branched structure, providing enhanced detection of molecular vibrations. We developed the RAMNet model, a convolutional neural network (CNN)-based architecture, to process hyperspectral Raman data and detect key molecular features. The AI-driven approach effectively improved noise reduction, spectral smoothing, and peak detection, enabling accurate identification of amino acids like leucine, cysteine, and serine. The RAMNet model demonstrated high accuracy in spectral classification, as confirmed by confusion matrix analysis and loss/accuracy curves. This work illustrates the potential of combining AI with SERS for precise molecular analysis, opening new avenues for applications in biochemistry, drug discovery, and environmental monitoring.</p>