How the index works.
How immortal.life automatically discovers, classifies, links, updates, and publishes longevity intelligence without human review.
What the system does
immortal.life retrieves public metadata from named literature databases, trial registries, Crossref-linked integrity records, and official regulatory feeds. Automated matching rules connect records to tracked topics, normalize dates and identifiers, and publish the resulting index.
What the system does not do
No scientist, clinician, researcher, editor, or human reviewer screens individual records before publication. The system does not reproduce a peer-review process, judge clinical suitability, or provide personal medical advice.
Evidence labels
Labels describe source metadata and study design signals: synthesis, randomized human study, human study, preclinical research, preprint, or general research record. They are navigation aids, not quality scores or treatment recommendations.
Topic matching and quarantine
Controlled terminology is evaluated across titles, abstracts, source keywords, and study types. Broad terms such as sleep, exercise, stem cells, or gene therapy also require explicit ageing, longevity, healthspan, or frailty context. Records scoring below 60% remain stored in automated quarantine and are excluded from public pages, counts, feeds, briefings, graphs, indexing notifications, and the sitemap.
Duplicate clustering
DOIs, registry identifiers, and normalized titles create deterministic clusters. Only the canonical cluster leader is eligible for publication; duplicates remain quarantined and traceable internally.
Public explanations
Published records display their relevance confidence, matched fields, controlled-term reasons, source-quality score, and freshness score. These explain automated routing and do not measure treatment effectiveness, safety, or overall scientific quality. Aggregate results appear on the quality telemetry page.
Freshness and provenance
Each record retains a direct source URL and update timestamps. Feed health is shown publicly. If live data cannot be retrieved, the interface does not substitute uncited claims.
Known limitations
Source metadata can be incomplete, delayed, duplicated, corrected, or wrong. Automated summaries may omit context. Registration does not establish trial quality, safety, effectiveness, completion, or approval.