For the patient sitting across from an endocrinologist with a suspected rare hormone disorder, the standard of care today looks something like this: a workup of labs and imaging, a referral chain that can stretch months, and a clinician who, between appointments, tries to keep up with a literature base that adds thousands of new PubMed entries every week. Specialty societies publish guidelines on a multi-year cadence. UpToDate and DynaMed remain the workhorse reference tools. For anything off the well-trodden path, including familial endocrinopathies, endocrine hypertension, or off-label hormone and peptide protocols that patients increasingly arrive asking about, the physician is largely on their own with a search bar.
Hormonaly, founded by endocrinologist and metabolic geneticist Dr. Fady Hannah-Shmouni, is betting that the answer is not another search interface but an evidence layer that does the reading first. The company describes itself as "the evidence OS for modern medicine," with 24 specialized agents that synthesize GRADE-rated clinical answers from more than 10,000 PubMed studies in under 100 milliseconds [Hormonaly.ai]. GRADE, the Grading of Recommendations Assessment, Development and Evaluation framework, is the same methodology that bodies like the WHO and the Endocrine Society use to weight evidence quality. Building it into the retrieval layer, rather than asking a clinician to apply it after the fact, is the architectural bet.
The wedge: hormone and performance therapeutics
The company's first consumer-visible product is Anabol.ai, an evidence platform covering steroids, peptides, SARMs, and regenerative medicine. It carries 240+ compound profiles with PubMed-sourced studies, side effects, drug interactions, and dosing protocols [anabol.ai]. This is a deliberate wedge. Hormone and performance therapeutics sit in a category where patient demand has run far ahead of clinician comfort and regulatory clarity. GLP-1 agonists, testosterone optimization, peptide therapies like BPC-157, and a long tail of compounds circulate through clinics, telehealth services, and gray-market channels with limited synthesized evidence available to the prescribing physician. The FDA has issued repeated warnings on compounded peptides and unapproved SARMs, and the clinician asked about them often has nowhere efficient to turn.
Hannah-Shmouni's framing, in podcast appearances, has been the shift from what he calls "sickcare" to "healthcare" [Leaders On Purpose Podcast]. Translated into product terms: meet clinicians and informed patients where the questions are actually being asked, in domains the traditional reference tools underserve, and earn trust through citation discipline rather than chatbot fluency.
Why the bet could be big
The clinical evidence-synthesis market has been dominated for two decades by Wolters Kluwer's UpToDate and EBSCO's DynaMed, both of which rely on human editorial teams updating monographs. Generative AI has cracked open a category that had calcified. Companies including OpenEvidence, Glass Health, and Atropos Health are all pursuing variants of the same idea: ground large language models in peer-reviewed literature and surface answers fast enough to fit inside a clinical encounter. OpenEvidence reached a reported $1B valuation in 2024 on the strength of free clinician access and rapid adoption.
Hormonaly's differentiation rests on two choices visible in the public product. The first is the GRADE rating layer, which forces the system to expose the strength of evidence behind each answer rather than presenting all citations as equivalent. The second is vertical depth in endocrinology and hormone therapeutics, a domain where the founder has authored over a hundred scientific publications and previously led research at the NIH [Mizter Rad Show]. In a category where general-purpose evidence tools are racing to cover every specialty shallowly, going deep in a high-question, high-liability vertical is a defensible posture.
The founder
Hannah-Shmouni's clinical credentials are central to the company's credibility. He is an internist, endocrinologist, and metabolic geneticist specializing in neuroendocrinology, rare endocrine genetic syndromes, and hypertension [ResearchGate; Elsevier]. He served as associate program director of the NIH's Inter-Institutes Endocrinology and Metabolism Fellowship Program and as principal investigator for endocrine genetic and hypertension disorders in the Stratakis Laboratory's section on endocrinology and genetics [Erdheim-Chester Disease Bio]. He sits on the editorial board of Elsevier's Endocrine and Metabolic Science. For an evidence-synthesis product whose entire value proposition rests on whether clinicians believe the synthesis, that résumé is the company's most important asset today.
| Hormonaly at a glance | Detail |
|---|---|
| Product | Evidence OS with 24 specialized AI agents |
| Literature base | 10,000+ PubMed studies |
| Response latency | Under 100 ms (company-stated) |
| Evidence framework | GRADE-rated answers |
| Consumer-facing app | Anabol.ai, 240+ compound profiles |
| Founder | Dr. Fady Hannah-Shmouni, MD FRCPC, formerly NIH |
What bears say, what bulls answer
The most credible concern is competitive. OpenEvidence, Glass Health, and Atropos have raised meaningful capital and are signing health-system deals, and the general-purpose frontier models from OpenAI and Anthropic keep getting better at medical question-answering on standardized benchmarks. A vertical evidence tool risks being squeezed between well-funded horizontal competitors above and free LLMs below. The bull answer, supported by the public product, is that hormone and performance therapeutics are precisely the kind of messy, off-guideline, high-stakes domain where horizontal tools underperform and where a specialist-built system with explicit evidence grading has a real edge. Anabol.ai also gives Hormonaly a direct channel to clinicians and informed patients in a category most competitors avoid for reputational reasons, which is itself a moat.
There is also a regulatory question worth flagging honestly. Clinical decision-support software in the United States sits in a nuanced FDA category under the 21st Century Cures Act, with carve-outs for tools that allow clinicians to independently review the basis of recommendations. An evidence-synthesis product that surfaces graded citations is well-positioned within that carve-out, but the line between "reference tool" and "regulated CDS" is one the company will need to manage as the agents take on more decision-weight.
What to watch
The next 12 months should answer three questions. Does Hormonaly raise an institutional seed or Series A round, and who leads it? Does Anabol.ai convert its current clinician and patient interest into a paying base that validates the wedge? And does the company publish, or partner on, any peer-reviewed validation of its synthesis quality against expert clinician benchmarks? That last milestone matters most. In evidence medicine, the tool that synthesizes the literature eventually has to submit itself to it.
Disease state: rare endocrine disorders and hormone and performance therapeutics. Patient population: adults navigating endocrinopathies, hypertension of endocrine origin, and the growing category of hormone, peptide, and regenerative medicine users currently underserved by mainstream clinical reference tools.
Pulse Raman, Health and Bio Correspondent