🤖 AgentIQ Live Dashboard

Real-time Agent Monitoring & Analytics Platform

🏥 System Health

Database Connected
LLM Evaluator Standby
API Server Online
Tables Ready 4/4

📊 Live Agent Metrics

Total Sessions --
Active Agents 3 Types
Success Rate 78%
Avg Response Time 1.2s

🎯 Agent Performance by Type

🧠 Intent Classification

📝 Live Activity Log

[15:45:32] Customer support agent processed billing dispute
[15:45:28] Sales agent qualified enterprise lead (50 users)
[15:45:25] Technical agent resolved API 500 error
[15:45:20] Intent classifier: billing_dispute (confidence: 0.94)
[15:45:18] Multi-agent workflow started: customer_support_session_123

🔧 How Developers Integrate

Single Line Integration:

agentiq.trace(session_id, "agent_response", user_input, agent_output, metadata)

Multi-Agent Workflow:

# Agent 1: Intent Classification agentiq.trace(session_id, "intent_classifier", user_query, "billing_dispute", {"confidence": 0.94}) # Agent 2: Billing Specialist agentiq.trace(session_id, "billing_agent", "investigate_charge", "found_duplicate", {"refund_amount": 29.99})

🌐 Live API Endpoints

🔍 Test Endpoints:
GET /health - System Status
GET /docs - Interactive API Documentation
GET /admin/database-status - Database Info
✅ Integration Ready: Send POST to /ingest/json with your agent data

🔄 Multi-Agent Workflow Example

📥 User Input

"I was charged twice for my subscription"

🧠 Intent Classifier

Route: billing_dispute
Confidence: 94%

✅ Logged to AgentIQ

💰 Billing Agent

Found duplicate charge
$29.99 on March 15

✅ Logged to AgentIQ

🔄 Refund Processor

Refund processed
ID: rf_xyz789

✅ Logged to AgentIQ

💬 Communication

Customer notified
Satisfaction: 89%

✅ Logged to AgentIQ
🎯 Result: Complete workflow visibility, performance metrics, and failure analysis