Health, explained.
Guides, insights, and plain-language explanations for anyone who wants to understand their health — and the AI helping them do it.
What We Say No To, and Why That Is the Point
In clinical software the refusals matter more than the features. Here are ours, and what each one costs us.
Why We Price Per Clinician, Not Per Query
A meter that moves every time someone uses the product teaches them to use it less. That is the opposite of what a clinic is buying.
Building AI Into What a Hospital Already Runs
Not another system to log into. The useful version lives inside the process rather than beside it.
For Manufacturers: Product Information That Holds in Every Language
Product and safety information has to say the same thing in every market you sell in. That is a consistency problem before it is a translation problem.
For Pharmacies: The Same Question, a Thousand Times a Day
What is this for. What will it do to me. Can I take it with the other one. Answering it well, in the customer’s language, is a staffing problem at scale.
For Laboratories: Explanation That Travels With the Result
Every report you issue is read by somebody who may not understand a word of it, usually days before they can ask anyone.
The Hours a Clinic Loses to Work That Is Not Medicine
Most of what exhausts a practice is not clinical. It is administration that exists because nothing else could do it.
What a Doctor’s Agent Actually Does Between Patients
Not a dashboard to check. Work that is already finished by the time you look at it.
Asking the Questions You Would Not Ask a Person
How worried should I be. Is this normal. The questions people carry away from an appointment because there was no room to ask them.
Answers in 98 Languages, and Why That Is the Hard Part
Health is difficult enough in your first language. Doing it properly in the other ninety-seven is most of the engineering.
One Account, Every Member of the Family
Most often it is not even your report. It is your mother’s, and you are the one in the corridor trying to make sense of it.
Following Your Health on the Days Nothing Happens
Health apps go quiet exactly when most people are well, which is most of the time. That is a design failure, not a user failure.
Checking a Meal From a Photograph, Home Cooking Included
Most food tracking fails because it was built around barcodes and a scale. Neither exists in the kitchens most of the world cooks in.
Looking Up a Medicine Without Being Sold One
You ask about one thing and get one answer. Nothing is suggested, nothing is curated, and no list is ever put in front of you.
Reading a Prescription, Medicine by Medicine
Handwriting you cannot read, names you have never seen, and an instruction that assumes you already know. Each line, explained separately.
How a Lab Report Is Read in Under a Minute
What happens between photographing a page and understanding every value on it — and why the range on your own document is where it starts.
What AI Cannot Replace in Medicine
The parts of a consultation that look like overhead are frequently the parts doing the clinical work.
Privacy in Health AI: The Questions to Ask Any Provider
Five questions that separate a company that has thought about your data from one that has written a policy about it.
The Next Wave of Health AI Is Agents, Not Chatbots
A chatbot waits to be asked. An agent does the work between the asking, which is where almost all the time in a clinic actually goes.
Health AI in Emerging Markets Is a Different Problem
The constraint is not a shortage of intelligence. It is that the doctor is scarce, the document arrives by WhatsApp, and nobody is coming to explain it.
What Regulators Mean by “Medical Device”, and Why It Decides Everything
The difference between health information and a regulated device is not tone or disclaimer. It is whether you make a determination about a named person.
The Trust Problem: Why People Do Not Believe Health AI Yet
Scepticism about health AI is rational. The way to answer it is not reassurance — it is showing your working.
Why Most Health AI Falls Apart Outside English
The benchmarks are in English, the training data is in English, and the people with the least access to health information are not.
What Health AI Is Actually Good At in 2026
Stripped of the claims, the useful capabilities are narrower than the headlines and more valuable than the sceptics allow.
In Health AI, Being Right Matters More Than Being Fast
A reading that takes half a minute and is correct beats an instant answer that is confidently wrong — but only if the wait is honest.
“Sourced, or Not Said” — What That Means in Practice
Where we state a clinical figure we name who published it and when. Where we cannot, we leave it out, and that rule deletes features.
Explain, Never Diagnose — The Line We Will Not Cross
A system that takes your values and returns a diagnosis is a regulated medical device. We built on the other side of that line deliberately.
A Question Asked Half in One Language and Half in Another
Real consultations are not conducted in clean English. Handling how people actually speak is the hard part most health tools quietly skip.
Why We Read the Range Your Laboratory Printed, and Never Our Own
We hold no reference ranges at all. It is the single most consequential decision in the product, and it costs us something every day.
The Hardest Thing in Health AI Is Knowing When to Say Nothing
Any system can produce an answer. The engineering problem is building one that reliably declines.
What It Means to Build a Model for Health, Not for Everything
Generality is a virtue almost everywhere in AI. In clinical documents it is the thing that has to be given up first.
Why a General Chatbot Is the Wrong Tool for a Lab Report
A general assistant is built to answer anything. A lab report is the one document where answering anything is exactly the wrong instinct.
How to Read Your Blood Test Results: A Complete Guide
Blood tests are the most common medical test in the world — yet most people receive their results without understanding what any of it means. This guide explains every major panel in plain language.
What Is AI Health Report Reading — And Should You Trust It?
AI tools that explain medical reports are growing fast. We break down exactly how they work, what they can and cannot do, and what questions to ask before relying on one.
Understanding Your Cholesterol Results: LDL, HDL, Triglycerides Explained
Your cholesterol panel comes back with four numbers and your doctor gives you three minutes to discuss it. Here is everything you need to know about what those numbers actually mean for your heart health.
How to Manage Your Entire Family's Health Records in One Place
Most families juggle reports, prescriptions, and test results across WhatsApp messages, paper folders, and email attachments. There is a better way — and it does not require a medical background.