We describe a compact computational spectroscopy platform optimized for molecular recognition using metal nanoparticle assays. The objective is motivated by the urgent need for low-cost, portable and high-throughput sensors for point-of-care (POC) clinical diagnostics. Nanoparticle based sensing has been successfully demonstrated for diagnosis and monitoring of infectious diseases, drug discovery, proteomics, and biological agent detection. Molecular binding on the nanoparticle surface is transuded into an optical signal by modification of the nanoparticle extinction spectrum (via a shift in Localized Surface Plasmon Resonance) or by modification of the molecular scattering spectrum (via Surface Enhanced Raman Scattering). Translating a nanoparticle -based molecular recognition system into a functional miniature hand-held biosensor requires spectrometer designs optimized to large area nanoparticle assays and integrated spectral filtering to improve the signal specificity. Large population sampling with small population sensitivity is essential to highly sensitive nanoparticle assay analysis. We describe a multimodal multiplex spectroscopy (MMS) platform that samples the spectral response of up to 106 populations of 10-100 nanoparticles in parallel. The advantages of MMS approach include: extremely high signal throughput due to its large aperture and high resolution with small form factor. We will demonstrate a nanoparticle biosensor platform based on MMS. Ultimately, a fully integrated functional miniature nanoparticle based biosensor for real time disease diagnosis in whole blood assays can be realized.