EnterpriseCommand Center

Enterprise AI Observability

Operational visibility from business use case to individual AI execution
refresh 30s · UTC−4
Active Use Cases
16
of 16 registered
AI Sessions
27.1K
24h · vs previous 24h
AI Spend
$155.3K
24h · vs previous 24h
Total Tokens
2.20B
vs previous period
P95 Session Latency
6.8s
vs previous period
Error Rate
2.75%
vs previous period
Quality Score
90.4%
vs previous period
Business Teams
3
vs previous period

What changed

AIHouse explains the biggest movements vs the previous period · click to investigate
Error rate 0.8% → 6.7%INC-2201 · Customer Support Copilot: OpenAI error rate increased from 0.8% to 6.7%Open incident P95 latency 3.7s → 5.2sINC-2198 · Branch Performance Assistant: P95 latency increased 42%Open incident Tokens / session 48K → 127KINC-2194 · Marketing Content Assistant: Token consumption anomalyOpen incident Quality score 93% → 88%INC-2187 · Proposal Generation Copilot: Quality score declined below thresholdOpen incident

AI Session Volume

sessions / 15m · stacked by platform · 24h
Σ 27,050
Amazon BedrockOpenAIAnthropicDatabricks

Token Consumption

tokens / 15m
2.20B total
InputOutputCached

Session Latency

P50 / P95 / P99 · seconds
alert ≥ 10s P95
P50P95P99

AI Spend Over Time

USD / 15m · stacked by cost type
Model/APIPlatform servicesInfrastructureFixed licenses

Session Health

sessions / 15m · share of 24h
SuccessWarningDegradedFailed
Success94.3%
Warning2.0%
Degraded0.9%
Failed2.7%

Cost vs Quality

bubble = session volume · click to drill in
■ high-value efficient zone■ high-cost / low-quality zone

AI Use Cases

click a row to investigate
16 shown
Use CaseTeamPlatformProviderModelSessionsTokensP95CostCost / SessionQualityHealth
Branch Performance AssistantSales & MarketingAmazon BedrockAnthropicClaude Sonnet2,41918.4M5.2s$1.8K$0.7696%Healthy
Customer Support CopilotCustomer OperationsOpenAIOpenAIGPT-56,842512.0M7.9s$38.4K$5.6287%Critical
Marketing Content AssistantSales & MarketingAnthropic APIAnthropicClaude Opus1,904241.0M9.4s$22.2K$11.6591%Warning
Data Analyst AgentFinance & AnalyticsDatabricks Model ServingMetaLlama 3.3 70B1,288164.0M11.2s$14.3K$11.0789%Warning
Internal Knowledge AssistantCustomer OperationsAmazon BedrockAmazonAmazon Nova Pro4,310286.0M4.1s$9.8K$2.2894%Healthy
Proposal Generation CopilotSales & MarketingAmazon BedrockAnthropicClaude Sonnet922198.0M8.6s$16.4K$17.8193%Healthy
Customer Opportunity AssistantSales & MarketingAmazon BedrockAnthropicClaude Sonnet64042.0M6.0s$3.1K$4.8892%Healthy
Finance Research AssistantFinance & AnalyticsAzure OpenAIOpenAIGPT-4.121431.0M6.6s$2.9K$13.3690%Warning
Release Notes GeneratorFinance & AnalyticsAnthropic APIAnthropicClaude Sonnet619.0M5.8s$640.00$10.4995%Healthy
Pricing AdvisorSales & MarketingAmazon BedrockAnthropicClaude Haiku58022.0M3.4s$690.00$1.1991%Healthy
Campaign Insights AgentSales & MarketingOpenAIOpenAIGPT-4.1 mini95096.0M6.1s$4.2K$4.4292%Healthy
Order Tracking AgentCustomer OperationsAmazon BedrockAmazonAmazon Nova Pro2,580150.0M3.9s$5.1K$1.9893%Healthy
Returns & Refunds AssistantCustomer OperationsOpenAIOpenAIGPT-52,390170.0M7.5s$13.1K$5.4885%Critical
Field Service CopilotCustomer OperationsDatabricks Model ServingMetaLlama 3.3 70B64080.0M10.4s$7.0K$10.9489%Warning
Invoice Processing AgentFinance & AnalyticsAzure OpenAIOpenAIGPT-4.154078.0M6.2s$7.1K$13.1591%Healthy
Demand Forecast AssistantFinance & AnalyticsDatabricks Model ServingMistralMistral Large77098.0M11.0s$8.5K$11.0487%Warning