SaMD pharmacovigilance
Software as a medical device breaks the assumptions of classical drug safety. A model that updates weekly does not have a stable exposure. We work on what surveillance means when the product changes under the patient.
Chenvar publishes on the problems it builds against. The research and the engineering are the same enquiry approached from two directions.
Software as a medical device breaks the assumptions of classical drug safety. A model that updates weekly does not have a stable exposure. We work on what surveillance means when the product changes under the patient.
Continuous biosensors generate signal that periodic reporting cannot represent. Our work on continuous glycaemic pharmacovigilance proposes a framework for it.
Algorithm-change control, clinical validation strategy and where the clinician sits in the loop when a model contributes to a decision.
A versioned corpus of guidance across EMA, MHRA, FDA and CDSCO/PvPI, structured so that a rule can be traced to the document and version that issued it.
Biosimilar surveillance, phenotype-specific drug risk in South Asian populations, and the pharmacovigilance gaps that follow from applying Western reference data.
Real-world evidence, literature surveillance and the evidentiary threshold at which a signal should change what a label says.
The full record is on ORCID, with new work added on a steady cadence.
Partnership, pilot, procurement or investment — enquiries are read and answered by the founder, not a queue.