Alert Processing Pipeline
How the agent transforms raw alerts into enriched tickets
Conviva
ITSM
SharePoint
Raw Conviva Payload
Incoming webhook data
{
"source": "Conviva",
"alert_source": "AI Alert",
"event_id": 1457888888,
"account_name": "c3.Parkin",
"root_cause": "iOS 26.1 and Riyadh",
"cumulative_impacted_unique_devices": 19192,
"metric_name": "Checkout Duration",
"alert_time": "02:45 Jan 18, 2026 PST",
"value": "21 sec",
"severity": "info",
"custom_fields": {
"additionalValue": "example"
}
}Agent Pipeline
Total: 4.2 seconds
Ingest Alert0.1s
Received Conviva webhook payload
Classify Severity0.8s
Reclassified from INFO to HIGH based on impact analysis
Search Knowledge Base0.6s
Semantic search across 847 articles
Search Historical Incidents0.4s
Pattern matching against 2,341 past incidents
Generate Diagnostic1.8s
LLM analysis with context from KB and historical data
Create ITSM Ticket0.3s
Enriched ticket with full diagnostic context
Route to Team0.2s
Auto-assigned based on root cause classification
Enriched ITSM Ticket
Auto-generated output
INC-5102HighAI Enriched
Checkout Duration Spike - iOS 26.1 Riyadh
TeamMobile App Support
SourceConviva #1457888888
Impacted Devices19,192
KB Articles2 matched
Historical Matches2 similar
AI Diagnostic Included
Full root cause analysis with correlation to historical incidents and recommended actions...
Before vs After
Value demonstration
Before (Manual L1)
- Raw alert with no context
- 30+ minutes investigation
- Manual KB search
- Inconsistent ticket quality
- Delayed team routing
After (With Agent)
- Enriched alert with full context
- 4.2 seconds processing
- Automatic KB matching
- Consistent, complete tickets
- Instant intelligent routing