Operator console
Understand cluster health, workload pressure, and the next safest action in seconds.
Cluster signal
Power trend from the two-second simulation loop
AI recommendation
Highest-priority demonstration decision
Live thermal, utilization, power, and allocation state
Decision model
Current operating policy
1. Sense
GPU telemetry + workload context
2. Score
Risk analysis + EcoScore
3. Decide
Safe placement + action
4. Explain
Operator-readable reasoning