Operational evidence, not investment performance. Every percentage shows its denominator; an unavailable or zero-sized sample stays N/A.
Loading measured reliability data...
ORION separates source availability from product correctness. A source being reachable does not prove that an entity was attributed correctly or that all matching trials were returned.
Large and small biotech companies, tickers, rare-disease assets, and co-developed drugs are rerun on a daily schedule. The cohort includes Leqembi, Enhertu, Casgevy, Applied Therapeutics, Cassava Sciences, Madrigal, and Epacadostat.
Resolved company, ticker, entity confidence, returned ClinicalTrials.gov count, terminal status, and error code are compared with the reviewed baseline. Attribution changes, confidence drops, trial-count collapses, and misleading source errors fail the check.
Leqembi/lecanemab was verified end to end as Eisai Inc. with 18 returned trials, including Clarity AD and AHEAD, and a confirmed saved report. A passing case is evidence for that path at that time, not a blanket accuracy claim.
Method details and limitations: release verification methodology. Operational incidents remain visible through the system status page.
Implementation, production connection, operational validation, and procurement readiness are different gates. This registry states the current public boundary without upgrading planned or pilot functionality to LIVE.
Machine-readable source: orion-capability-manifest.json
Review the methodology and limitations, evidence quality, coverage boundaries, change history, system status, privacy policy, terms, and guided pilot boundaries.