Journal ·
Why we're building Qaytona
Image by Osamu Yokonami
Qaytona is a clinical reference for practitioners who recommend supplements. Each ingredient has one monograph: the clinical evidence, the doses studied, safety and interactions, and how the forms on the market compare, with every claim linked to its source. We're building it because patients now learn about supplements faster than anyone can check them, and because an AI answer is only as good as the question behind it.
Patients arrive already taking it
Most people hear about a supplement online or from a friend long before they mention it to a practitioner. By the appointment they're often taking several, some the practitioner has never had reason to look up. The details that matter, like the form, the dose and the interactions, are rarely on the label.
Peptides are the clearest case right now. BPC-157 is promoted on TikTok and LinkedIn and by influencers selling their own versions. A 2025 review of its research found 35 animal and lab studies and one human study of 12 people, with no human safety data. It isn't an approved drug. 1
Patients ask about it anyway, and there are few neutral places to look.
You can't ask for what you don't know
AI has changed how people look things up. You ask a question and get a fluent answer. But the answer depends on the question, and the question depends on what you already know. A practitioner who knows that magnesium oxide has its own sleep trial asks a better question than one who assumes only glycinate works. With AI, what you bring to it matters more than ever.
That's the case for a library. You come for one thing and find what sits next to it: the interaction you didn't think to check, the form the trial actually used, the study that disagrees. A well-made reference turns up the questions you didn't know to ask.
How the library is built
Each monograph draws on two kinds of evidence. The academic side comes from PubMed, the NIH, the FDA and ClinicalTrials.gov, including trials still running. The real-world side is what patients and practitioners say on Reddit, X, YouTube and TikTok, kept apart from the clinical evidence and labelled as what people report.
Language models help us read and draft. Every citation has to point to a real paper on the claim's topic, safety, interaction and clinical evidence claims are each read against their paper, and a claim the paper doesn't support is removed. The methodology article shows how each monograph is made and checked.
The goal is a well-informed opinion: knowing what you should know, finding it in the minutes between patients, and going as deep as you want when you need to. The decision about a patient stays with you.
Who's building it
Dr. Eric Steiner is co-founder and chief medical officer. He started in medicine, took an MBA, built businesses with his father, then turned to health technology. He is the CEO of Mountain States Health Products, a Colorado company that has supplied practitioners since 1990. MHP sells supplements. Qaytona doesn't, and it carries no affiliate links or ads.
Parth Patel leads product and engineering. He's a designer turned full-stack builder who has led product teams, and runs Audemars De Luca, an independent design practice in New York whose clients include Google, Equinox and University of Michigan Health. More at
withparth.com.
What's next
Qaytona is a web app today, free while it's in beta. Mac and iPhone apps are in progress, along with a way to ask Qaytona from inside Claude and ChatGPT. What won't change is how the information is laid out, so the answer and the evidence behind it are always one step apart.
Qaytona is an educational reference and doesn't give medical advice.