Network Status
Overall OPERATIONAL
Active Alerts 3
Selected GW GW-07
Gedung B, Floor 3
Devices 142/150
94.7% online
AI Model
Base Model GPT-4
Fine-tuned IoT-Net v2.1
Response 1.2s avg
Context 128k tokens
Mode Live
Temp 0.3
History
Current Session
AI Agent troubleshooting...
Just now
Gateway Diagnostics
GW-07 packet loss analysis
Today, 14:22
Security Scan
Ran full security audit...
Today, 11:05
Routing Optimization
BGP route recalculation
Yesterday, 16:40
Device Onboarding
Added 12 new sensors...
Yesterday, 09:15
Network Health Check
Weekly health report gen...
2 days ago
Topology Mapping
Full topology scan done
3 days ago
AI Network Agent
LLM-IoT-Agent v2.1  ·  GPT-4 backbone
Ready
─── Session started · Secure channel established ───
Model Info
LLM-IoT-Agent v2.1
Confidence Score 94%
Queries today 0
Avg latency —
Tokens used 0
API endpoint api.openai.com
Status Healthy
API Call Log LIVE
Reasoning Trace

1.Load network topology graph from cache (nodes: 150, edges: 312)
2.Retrieve last 24h telemetry for selected gateway GW-07
3.Index alert history — 3 unresolved critical alerts found

1.Parse user intent → classify as [DIAGNOSTIC] query
2.Select tool: GET /api/v1/gateways/{id}/status
3.Cross-reference with baseline KPIs (latency, packet loss, CPU)
4.Generate structured diagnostic report with root-cause analysis
5.Rank recommendations by impact × effort matrix

1.Verify response against network policy constraints
2.Check recommendation safety — no disruptive changes flagged
3.Confidence threshold met (≥ 80%) — response approved