<p>This study presents the development and characterization of a metasurface-based sensor for high-precision peptide detection. The architectural design that constitutes the sensor design incorporates a hybrid nano-photonic structure, featuring engineered square resonator with optimized geometric parameters. Rigorous electromagnetic analyses were performed utilizing finite element method (FEM) simulations in COMSOL Multiphysics to demonstrate the sensor’s electromagnetic response characteristics. Quantitative performance metrics were systematically evaluated, including quality factor optimization, sensitivity measurements, and figure of merit (FOM) calculations. The proposed sensor demonstrates exceptional sensing capabilities, achieving sensitivity values of 500 GHzRIU⁻¹. Spectroscopic resolution was validated through full width at half maximum (FWHM) measurements of 155&#xa0;GHz. Performance optimization was further enhanced through the implementation of machine learning algorithms essentially random forest regression (RFR), for automated pattern recognition and response prediction. The resultant regressor models, trained across diverse parametric configurations, exhibited superior predictive accuracy with coefficient of determination (<i>R</i>²) values approximating unity across all the test cases. These results demonstrates the efficacy of the proposed sensor for label-free, non-invasive biomolecular detection and characterization, with significant potential for implementation in clinical diagnostics and diverse biomedical applications.</p>