Company

New medicines, proven on virtual patients first

We are building the simulation layer every drug programme runs before, and alongside, its physical trials, so development gets faster, safer and fairer to the people trials leave out.

Why we started

Too many trials fail on choices made too early

Clinical trials are the slowest and most expensive step in drug development, and most fail. Often the cause is a choice made before the first patient enrolled: a dose, an inclusion criterion, an endpoint.

Those choices are made with thin, biased history. Groups that matter, from children to under-represented ancestries, are often missing entirely. We started Sanenza to let teams test those choices on realistic virtual populations first, and to keep an honest record of how well the simulations predicted what happened next.

Principles

How we work

Evidence

Check every claim

A prediction is a claim until it is compared with an outcome.

Equity

Simulate the people trials miss

Rare and under-represented groups are a design input, not an afterthought.

Restraint

Say what we cannot do

Our reports state their limits in plain words.

Roadmap

Draft roadmap

Plans, not commitments. Dates depend on design partners.

WhenMilestone
2026Platform core: Cohorta, Insilon and Evidra in development
2027 H1First design-partner pilots in one or two therapeutic areas
2027 H2Protova and Vigila in pilots; private cloud deployment
2028Physova pilots; wider therapeutic-area coverage

Careers

Open roles

Draft openings. Write to careers@sanenza.com with the role in the subject line.

Scientific modellingLead QSP and PK-PD scientistHybrid, Boston or London
Machine learningSenior ML engineer, physics-ML surrogatesRemote, US or EU
Machine learningML engineer, generative cohort modelsRemote, US or EU
PlatformGPU platform engineerRemote, US or EU
Regulatory scienceRegulatory science lead, MIDDHybrid, Boston or London

Stack

What we build with

NVIDIA BioNeMoBiomolecular target, ligand and pathway response
NVIDIA ModulusPhysics-ML surrogates for physiology and PK-PD
NVIDIA ParabricksGPU genomics that grounds patients in real omic variability
RAPIDS cuDF, cuML, cuGraphEHR, omics and outcome-graph processing
Clara and MONAIImaging-derived phenotypes and anatomy
NeMo and GuardrailsTrial-design agents and evidence drafting
Python and JAXModelling and research code
KubernetesGPU job orchestration
PostgreSQLLedger and metadata

Help prove medicines on virtual patients first

If you work on pharmacology, machine learning or regulatory science, we would like to hear from you.