Programme evidence spine
Select or create a programme. Every linked module will inherit the same programme identity.
Discovery Programme OS · Beta
One programme.
One programme.
Every scientific decision.
Connect targets, registered candidates, approved assays, exact model versions, benchmark runs, experiments, module handoffs, evidence gaps, reviewer approvals, and change impact without pretending that virtual evidence is experimental proof.
Secure mutation access
Read-only programme views may be public. Writes use the backend service-token boundary. Never commit credentials into source code.
Decision Pack
Compile current candidates, assays, exact model lineage, experiments, handoffs, decisions, patent questions, outcomes, impact alerts, and unresolved gaps.
Not compiled.
Scientific source graph
Trace programme → candidate → assay/model/experiment → handoff → decision relationships.
Not generated.
Next best experiment
Transparent value-of-information ranking. The score prioritizes review; it never authorizes an experiment.
No ranking loaded.
Create discovery programme
Ready.
Programme list
Register candidate
Ready.
Candidate registry
Assay Data Gateway
Only approved, quality-complete records become training-eligible.
Governed active-learning campaign
Campaigns are blocked when approved assay evidence is absent.
Assay records
Link exact model use
Missing checkpoint, commit, container, or context fields remain visible as lineage gaps.
Register benchmark run
A benchmark is incomplete until dataset identity, leakage controls, seeds, checkpoint, code, and uncertainty are recorded.
Model lineage
Plan decision-changing experiment
Ready.
Ranked experiment portfolio
No ranking loaded.
Experiment plans
Transactional module handoff
A handoff is a durable transaction, not a copy-and-paste instruction.
Handoff history
Record reviewed decision
Ready.
Decision history
Pocket hypothesis
A predicted cavity is not automatically a druggable pocket.
Patent & FTO review queue
The system never issues a final patentability or FTO conclusion.
Watchtower impact
Licence and security changes trigger mandatory review.
Discovery portfolio cockpit
The portfolio shows evidence completeness, unresolved risks, and next decisions. It does not fabricate a probability of technical success.
Not loaded.
Product capabilities and honest maturity
Operational, Beta, Prototype, Research, Roadmap, and External integration are evidence labels—not marketing scores.