Stanford and Pumpkinseed use single-cell Raman and ML to detect bacterial pathogens in wastewater. This proof-of-concept work establishes that Raman can be used to detect biothreats without amplification for wastewater-based epidemiology.

Collaborating with Stanford, Pumpkinseed co-founders Nhat Vu and Jennifer Dionne demonstrate that single-cell Raman spectroscopy and AI can predict melanoma treatment resistance with 91% accuracy across patient-drug combinations.

Pumpkinseed co-founder Jack Hu, intern Varun Dolia, and Stanford Professor Jennifer Dionne are co-authors on new research introducing DynaSpec, a computational framework that maps the relationship between molecular structure and Raman spectral signatures — achieving over 85% classification accuracy across 13 closely related glycan structures. The work advances the scientific foundation underlying deSIPHR's ability to read complex biomolecules directly from their vibrational fingerprints.

Pumpkinseed Bio's nanophotonic platform is unlocking a new era in precision cancer immunotherapy — enabling label-free, AI-powered tumor profiling at the single-cell level.

Pumpkinseed and Stanford researchers have built a silicon nanoantenna array that breaks a longstanding optical trade-off — enabling ultra-precise, chip-scale light control for next-gen sensing, imaging, and computing.

Pumpkinseed co-founder Jack interfaces high-Q resonators with molecules to detect COVID-19 genetic fragments in minutes — no amplification needed. This label-free platform could redefine rapid, compact diagnostics for disease and beyond.

Pumpkinseed and Stanford researchers used an acoustic bioprinter to split complex fluid samples into millions of micro-droplets, then used AI-powered Raman spectroscopy to identify pathogens with 99% accuracy — even in blood.