Live federated learning analytics — tenant nodes, promoted patterns, shadow rejects, and contribution breakdowns from the real backend.
Local candidate patterns are generated privately, confirmed through votes, and promoted back into the detector when confidence is high.
New local proposals wait in shadow before they influence routing.
0 pending candidates
Only high-confidence, cross-tenant signals enter the shared detector layer.
0 promoted patterns live
Weak or stale patterns remain visible for auditability and future review.
0 shadow rejected · 0 retired
Live data from the federated pattern repository.