Some of these will take a decade. Some may never work. We share them anyway — because closing the diagnostic gap for four billion people demands the audacity to attempt what the field still calls impossible.
Why we publish our moonshots
Every programme on this page is research — not a product, not a promise. We label the stage on each, and we are candid that frontier science fails often and slowly.
But the largest problems in diagnostics will not be solved by incremental thinking. So we aim far past today's instruments — at instant answers, zero reagents, intelligence on every device, and a working lab in places that have never had one. We state these goals out loud so the right scientists, partners, and institutions can help us reach them.
The moonshots
Each is a long-horizon research direction at concept or early stage. None is available, validated, or proven. Together they describe the world we are building toward.
"Any test, from a single drop — with zero reagents."
One reconfigurable carrier that runs any class of assay by reading the intrinsic physics of a sample — optical, electrochemical, thermal — and letting AI, not chemistry, do the interpreting. The end of one-test-one-kit.
Frontier research"A full blood panel in under a second."
Reading the complete spectral fingerprint of a sample with light alone — no incubation, no waiting, no perishable chemistry — so a result is limited only by the speed of computation, not chemical reaction time.
Open research question"A lab-grade result with no technician — ever."
Instruments that calibrate, run quality control, and validate themselves against reference physics and on-device AI — removing the single biggest barrier to decentralised testing: the need for trained hands and a central lab to trust the number.
Concept"Catch disease years before the first sign."
Continuous, ambient, non-invasive biosignals — from wearables, breath, saliva — interpreted by a model that learns the earliest physical signatures of disease, shifting medicine from reaction to genuine prediction.
Long-horizon ambition"A diagnostic node anywhere on Earth — no grid, no network, no cold chain."
Solar-powered, offline-first, reagent-free diagnostic nodes rugged enough to deploy in the hardest places on the planet — so the nearest lab is never more than a walk away, for anyone.
Frontier research"One intelligence that can read any biological signal."
A multi-modal foundation model for human biology — an interpreter for raw biosignals the way a language model interprets text — small enough to run on the device, at the patient, with no cloud dependency.
Frontier research"A real-time weather forecast for disease."
A mesh of decentralised diagnostic nodes feeding anonymised, privacy-preserving signals into a live map of outbreaks and antimicrobial resistance — turning millions of individual tests into a planetary immune system.
Long-horizon ambition"Sensing chemistries with no cold chain, no expiry, no waste."
Engineering biomolecules and solid-state sensing surfaces that are room-temperature stable, self-amplifying, and fully biodegradable — or removing perishable chemistry from the diagnostic path altogether.
Open research questionPlus a growing atlas of open questions in biophysics, microfluidics & reagent science
Honest by design
We publish these because audacity, stated transparently, is how hard science attracts the people and partners who can advance it. Sharing the goal is not the same as claiming the result.
None of the programmes above is a commercial product, a regulatory-cleared device, or a clinically validated claim. Each will be held to the same evidence gates as everything else we build — concept, prototype, analytical verification, clinical validation, regulatory pathway — and we will report what works and what does not.
Frontier research and concept stage. Forward-looking ambitions, not available products or validated performance claims.