12 Production & Testing Quick Scenarios
Click any scenario to auto-populate operational settings and execute validation
Traffic Control & Rollout Policy
Adjust traffic allocation between the stable version (Baseline) and candidate (Canary)
BLUE (Baseline)90%
gpt-4.1-mini
GREEN (Canary)10%
gpt-4.1-mini
100% Blue50/50100% Green
Traffic Injection & Chaos Engineering
Simulate production workload and inject faults to validate system resilience
Batch Size (Requests):
Observability & SLO Dashboard
Consolidated real-time metrics, latency percentiles, and Error Budget consumption
Total reqs:0
P95 Latency (SLO < 2.5s)
0ms
SLO HEALTHYalvo: <2500ms
Average TTFT (Time-to-First-Token)
0ms
P50 Mediana: 0ms
Error Budget Status
100%restante
Circuit Breaker State
CLOSED
Estimated Cost (USD):$0
Distribuição de Latência (Blue vs Green)
Comparação percentilar em tempo real com linha de referência de SLO
Blue (gpt-4o)
Green (gpt-4o-mini)
P50 (Mediana)
0ms|0ms
P90
0ms|0ms
P95 (SLO Gate)
0ms|0ms
P99 (Worst Case)
0ms|0ms
CUSTO BLUE (gpt-4o)
$0
0 tokens
CUSTO GREEN (gpt-4o-mini)
$0
0 tokens
Real-Time Invocation Trace
Detailed audit trail of each routed inference request with status and tokens
No routed requests yet. Pick a scenario or trigger a simulation above.