About HypeCheck
What is HypeCheck?
HypeCheck helps you go from claim to evidence. When a supplement claims "this helps you lose weight" or "supports gut health," we trace the chain back: which ingredient is that claim based on, what study supports it, at what dose, and does the product actually contain that dose?
We do this whether the answer is "yes, it checks out" or "no, the claim doesn't survive the dose math." We're calibrated, not skeptical — equal-opportunity transparency for legit products and overhyped ones.
Who's behind this
Danny Kim
Founder & Editor
PhD Computer Engineering, University of Maryland (2018)
Day job: Systems Engineer at Endor Labs, where I build large-scale code-analysis systems used to find vulnerabilities in open-source software.
I'm an engineer, not a doctor or dietitian. My PhD work focused on detecting malware by analyzing low-level execution patterns at scale — essentially, separating real signal from marketing-style noise across millions of programs. HypeCheck applies that same lens to supplement marketing: trace the chain from claim to evidence, ignore the rhetoric, judge based on the data.
I built HypeCheck because I kept almost-buying supplements off Instagram and realized there was no fast, neutral way to check whether the claims held up. The site is not paid by any supplement brand. There's no monetization today; transparent affiliate disclosures or a paid tier may come later.
What I'm not: a medical professional. HypeCheck is an evidence audit tool, not medical advice. If you have a condition, take medications, are pregnant, or have liver/kidney concerns, talk to a clinician before changing anything.
How analyses are produced and reviewed
Every analysis on HypeCheck is generated by large language models (Claude Sonnet and GPT-4) running against retrieved PubMed research, NIH databases, and a structured ingredient knowledge base we maintain. The LLMs do not invent citations — every PMID gets verified against the actual retrieval set before publishing.
I (Danny) spot-check analyses for obvious issues before they're public, and review any flagged analyses in detail. Analyses are not individually reviewed by a credentialed dietitian or MD — that's a Phase 2 commitment as the site grows. If you spot something wrong on a review, the Report Issue button at the bottom of every page goes straight to me.
The problem we're addressing
The supplement industry is a $60+ billion market with minimal regulation. Consumers see health-product ads on TikTok, Instagram, and podcasts, but have no fast, neutral way to verify whether the claims are backed by science or whether the doses actually match clinical evidence. Googling "is [product] a scam?" turns up affiliate-driven fake reviews or forum posts without citations. HypeCheck is the tool we couldn't find.
Our Approach
- Calibrated: Equal-opportunity transparency. We highlight what works as carefully as we debunk what doesn't. Verdicts are downstream of evidence, not the headline.
- Evidence-based: Every analysis cross-references ingredients against PubMed research, NIH databases, and clinical dosing guidelines.
- Transparent: We show our work. Every ingredient gets an evidence rating, effective dose comparison, and source citations.
- Independent: Not paid by any supplement company. Analyses are AI-generated (Claude Sonnet + GPT-4) and spot-checked by the editor before publishing.
- Fast: Traditional lab testing takes weeks. Our AI analysis can evaluate a product within minutes of receiving a URL.
Important Disclaimers
- HypeCheck is an informational tool, not medical advice. Always consult a healthcare professional before starting any supplement.
- Our analyses evaluate claims and ingredient labels as published by the manufacturer. We do not perform lab testing of physical products.
- AI-generated analyses are reviewed for accuracy but may contain errors. We encourage users to verify findings with their own research.
- We never use language like "scam" or "fraud" about specific companies. Our ratings reflect the evidence behind product claims.
Want to learn how our analysis engine works? Read our Methodology page.
Questions or feedback? See our Terms of Service or Privacy Policy.