Evidence intelligence,
not content.
Therapeutic peptides are the most over-hyped, under-explained category in modern health. The story is told in two registers — bro-science on one side, paywalled trials on the other — and almost no one is translating between them in good faith.
Mydosage exists for the people in the middle. The literate, curious, non-clinician reader who wants to know what a peptide does, what the evidence actually says, and which claims have outrun their data. We are not selling vials. We are not running an affiliate program. We are not optimising for time-on-page.
We are building the reference layer this category never had: an operating system for therapeutic intelligence, calibrated against human studies, surfaced with uncertainty intact.
- 01
Evidence intelligence.
Not content. Not commentary. A reference layer that synthesises the evidence and admits its uncertainty in the same gesture.
- 02
Calibrated, not certain.
Every score on this site has a confidence interval behind it. Every claim has a citation behind it. Nothing is asserted that the evidence cannot carry.
- 03
Suggestive, never prescriptive.
Mydosage surfaces what the human data shows. It does not tell you what to take. The clinician is in the loop. Always.
- 04
Reviewed quarterly.
The science moves. The index moves with it. Every entry carries the reviewer's byline and the date of last revision. The changelog is public.
What it is
- — An independent reference layer.
- — A scoring system that re-derives itself.
- — A scientific consensus engine.
- — A translation layer between trials and decisions.
- — A surface that admits what it doesn't know.
What it isn't
- — A clinic. We do not prescribe.
- — A vendor. We do not sell peptides.
- — A community. We do not run a forum.
- — A chatbot. The AI is bound to indexed evidence.
- — A SaaS dashboard. There is no funnel.
- Indexed citations only. Every claim is bound to a citation row in the index. The AI cannot synthesise beyond the index.
- Calibrated scoring. The Evidence Intelligence Score™ is deterministic — it derives from human-trial weight, mechanism specificity, follow-up depth, and study consistency. It is not hand-tuned per peptide.
- Quarterly review. Every entry is re-examined at least once per quarter. When evidence moves, the score moves, and the change is logged publicly in the changelog.
- Reviewer byline. Every published surface carries the reviewer responsible for it and the date of last review.
- Errata are first-class. When we are wrong, the correction is visible — not buried.