The clinical AI that answers for where you practice.
Primary AI answers from your guidelines, in your units, and tells you when the international evidence diverges. Built for the busy physician who needs a quick, cited answer at the point of care, not a page of studies to wade through.
Our founder
Primary AI is led by a practicing clinician trained at the University of Toronto and the University of British Columbia, with research experience at SickKids Hospital.
That clinical background shapes every decision we make: how answers are structured, what gets cited, what gets flagged as a safety risk, and what we refuse to do. We will not ship a feature that we wouldn't trust at the bedside.
Milestones
Terms of Service & Privacy Policy published
Published our professional-use Terms and a data-minimization Privacy Policy covering no-PHI handling, encryption in transit, and named AI inference providers.
Canadian guideline coverage added
Layered in Canadian national and society sources — CADTH, the CMA CPG Infobase, NACI, Health Canada, the Canadian Cardiovascular Society, Diabetes Canada, Hypertension Canada, the Canadian Task Force on Preventive Health Care, and Choosing Wisely Canada — alongside US and international evidence.
Clinical reference library launched
Opened a growing, publicly indexable library of evidence-based clinical reference pages at /q, each with verified citations and structured data.
Primary AI public beta
Primary AI opened to licensed clinicians — free, no credit card, no patient data required — with citation-verified answers grounded in current guidelines and landmark trials.
Public medical content with direct citation links
Primary AI searches publicly accessible medical content and public guidance. Each answer links to the cited public page when one is available. It also identifies differences between locations when they affect care.
How we keep answers safe
Hallucinated citations are the biggest failure mode of medical AI. We engineer specifically against it. Read the full detail on our Security & Trust page.
Private Assistant history
Clinical questions must be de-identified by default. Authorized clinicians can use patient information in Primary AI Assistant. Approved Canadian Business API customers can use a separate fail-closed PHI path. Assistant transcripts and notes are private to the user and deleted after 30 days.
Citations verified against live lookups
Every reference is checked against a live PubMed and DOI lookup so the paper actually exists before it is shown to you — the single biggest failure mode of medical AI, engineered against directly.
No-fabrication guardrail
A hard guardrail instructs the model to refuse rather than invent a citation when it cannot find a real one. We would rather say 'no source' than show you a fake one.
Evidence grading
Recommendations flag the type of evidence behind them — RCT, meta-analysis, guideline, or expert consensus — so you can weigh them at a glance.
Frequently asked questions
The questions clinicians ask most before they trust an answer.
Ready to ask your first question?
Primary AI is free for licensed clinicians. No credit card, no patient data required.
Start asking