About Alter Health AI

What we do

Alter Health AI builds evidence-based evaluation tools for personal health products. Our software analyzes supplement labels, cosmetic ingredients, and product claims using publicly available scientific information, then presents findings in a way that helps consumers make informed decisions.

Our approach

We aggregate and structure information from peer-reviewed research, official guidelines, and reputable health authorities. We do not manufacture or sell products. We do not generate clinical conclusions of our own — we present what existing published evidence and guidelines say, with citations, in a structured format.

Why this matters

The global supplement market exceeds $170 billion annually. Yet outside a handful of Western markets served by services like ConsumerLab and Labdoor, users have no independent way to evaluate what's in the products they buy. Pharmacy chains recommend products they sell. Marketplace reviews are unreliable. Manufacturer claims are not independently verified.

We focus on the markets where this evidence gap is widest — regions where consumers have no independent, native-language way to evaluate what's in the products they buy. Hundreds of millions of people worldwide lack access to evidence-based supplement evaluation tools, and closing that gap is why we exist.

Our standards

We cite sources for every claim. We treat "unknown" as a valid assessment outcome rather than fabricate confidence. Clinical reviewers approve content before publication. We do not provide medical advice; users with specific health conditions should consult licensed healthcare professionals.

How we handle AI reliability

Our products use artificial intelligence to extract product information from labels, match products to evidence sources, and present findings. AI systems can produce confident-sounding output that is factually incorrect — a failure mode commonly called hallucination. We address this directly:

No clinical claim is surfaced without a verifiable source. Every evidence statement in our products is linked to a specific PubMed citation, regulatory document, or named scientific authority. If we cannot cite a source, we do not make the claim.
"Unknown" is a valid result. When evidence is insufficient, contradictory, or absent, our products say so plainly rather than generating plausible-sounding fabrication.
Curation happens offline, separately from runtime. The evidence database that backs our assessments is curated by our team reviewing primary literature. AI is used to extract and present this curated information at runtime, not to generate medical claims on the fly.
Human review before publication. Alter Health AI LLC publishes evidence assessments. Clinical reviewers approve content before publication. Reviewer credentials available on request.