Clinical layer
Study records are indexed by NCT ID and normalized into sponsor, status, phase, disease and country fields. Current production contains 5,964 unique indexed NCT IDs.
Methodology
Scores are signals,
not claims of certainty.
NeuroRisk.ai combines source records with derived analytical layers. Every score should be interpreted in context, with freshness, coverage and data limitations visible.
Research layer
PubMed/NCBI and OpenAlex-derived intelligence contribute publication, author and institution signals. Source-specific freshness is tracked independently.
Market layer
Country profiles combine clinical, research and derived opportunity signals. Current priority-market models expose opportunity, ML priority and confidence scores.
AI layer
Public AI Analyst Beta is deterministic over indexed data. Enterprise AI workflows must retain evidence, timestamps, confidence and provenance.
Quality gates
Unique clinical trial keys · null-key checks · source refresh logs · generated-page thresholds · public-page metadata QA · health checks after deployment.
Known limitations
Indexed counts are corpus counts, not guaranteed exhaustive global totals. Opportunity scores are comparative decision-support signals and are not clinical, regulatory or investment predictions.