We build the tools we
wanted to exist
cosys.work makes software for people managing long-term conditions and the clinicians who treat them. We started because the available tools asked for a permanent internet connection, a monthly fee, and an unconditional copy of the most private thing a person has.
Built in Nepal. The Diabetes App is in closed pilot and free forever.
Diabetes care is mostly logging. Readings, meals, doses and symptoms, recorded by hand several times a day, for the rest of someone’s life. It is tedious enough that most people stop, and when the logging stops the clinician loses the only honest view of what is actually happening between appointments.
The usual answer is more hardware and more cloud: a continuous monitor, a subscription, a dashboard hosted somewhere far away. That answer excludes most of the people who need it most, and it quietly assumes that a company should hold everybody’s medical history in one place.
We took a different route. Use the camera people already own. Run the intelligence on the phone. Move records directly between the patient and the people who need them, and let nothing in the middle read them.
That turned out to require three things rather than one: an app for the patient, an app for the clinic, and a peer-to-peer network connecting them. The patient app and the network come first; the Clinic App is in development.
Four commitments that make
the rest of it harder
The record belongs to the patient
Not to us, and not to whoever acquires us later. If a company can monetise a health history, sooner or later it is asked to. We removed the option by removing the copy.
Connectivity is a luxury, not a premise
Software that stops working when the signal drops is not a medical tool, it is a website. Everything essential runs on the device, and the network is a bonus.
Clinicians decide, software remembers
No dose calculations and no automated diagnosis. We build record-keeping and communication, and the clinician makes the call.
The people who live with it should steer it
Our diabetes trainers, one of whom lives with Type 1, shape what we build.
Built for the hard case first,
not retrofitted to it
We work out of Nepal, and we build for it before we build for anywhere else. That is a design constraint as much as a market.
Start where the constraints are real
We build first for places with patchy bandwidth, expensive hardware and long journeys to clinics. Software that holds up there works comfortably everywhere else, and the reverse is almost never true.
Meet people at the hardware they own
Continuous monitors and connected devices are out of reach for most people with diabetes. So the camera is the sensor: photograph the meter you already have, and let the phone do the reading.
Turn the journey around
Routine chronic care sends the patient to the clinic, the lab and the pharmacy. Where licensed local services can do it, we route the visit in the other direction instead.
Generalise from one condition, carefully
Diabetes is where we start because the daily logging burden is severe and the clinical loop is tight. The same fabric extends to other long-term conditions once it is genuinely proven on one.
What we make
Three products, one peer-to-peer fabric.
The Diabetes App
An offline-first diabetes companion, in closed pilot, free forever. Photograph your glucometer and lab reports; the number is read on your phone and you confirm it. Also called tda by cosys.
Learn moreThe Clinic App
In development, preview on request: a companion app for clinics. The history a patient chooses to share arrives over an encrypted connection before the consultation starts.
Learn moreThe P2P Platform
The peer-to-peer network underneath both apps: device-to-device sync over iroh, a mailbox that holds only sealed envelopes, and verified clinic nodes.
Learn moreWork with us
Clinics willing to shape the workflow, people living with diabetes who have strong opinions, and developers who want to build on a peer-to-peer network are all welcome.