SLO Thresholds & Error Budget Burn Rate

Configure P95/P99 latency limits, TTFT guarantees, and monitor real-time error budget consumption.

Active SLO Policies

3 monitored targets

Healthy Error Budgets

Breached / At Risk

SLO Target Name	Service	Metric Category	SLO Threshold	Current Value	Remaining Budget	Burn Rate	Status
Streaming TTFT (Time-To-First-Token) Guarantee slo-ttft-copilot	copilot-agent	time to first token	< 450 ms	382 ms	88.4%	0.8x normal	healthy
RAG Hybrid Vector P99 Latency SLA slo-p99-rag-dense	rag-service	p99 latency	< 850 ms	920 ms	14.2%	3.4x normal	breached
Institutional SEC Extraction Hallucination Ceiling slo-hallucination-cap	summarizer-agent	hallucination rate	< 4 %	3.8 %	32%	1.9x normal	at_risk