Clinical Development Command Center
One governed clinical-development record over the same BayesPharma programme and candidate identity. Evidence changes programme state; an AI explanation never does.
Development path
A phase is never advanced by a percentage. All required gates must be satisfied and an accountable human must approve the next transition.
1 · Programme & phase status
Clinical development extends the nominated Discovery Programme candidate; it does not clone the asset into another identity system.
2 · Study & analysis portfolio
Study plans are workflow objects, not authorization to dose participants. External protocol, ethics, regulatory and site controls remain authoritative.
Plan a governed study/work package
3 · Clinical evidence matrix
Only human-QC-approved observation imports become MEASURED. Simulations and models remain COMPUTED; registry/publication evidence remains RETRIEVED; accountable decisions remain REVIEWED.
4 · Deterministic clinical blockers
Blockers are derived from explicit evidence and review gates for the current phase, not from Qwen or an opaque readiness score.
5 · Next Best Clinical Work
Ranking: decision criticality × uncertainty × discriminatory power × probability of changing the decision ÷ cost/time burden. Every recommendation still requires human approval.
Human phase transition review
Only the next governed phase can be approved. A blocked gate prevents approval server-side.