Full Observability
for LLM Apps.
Real-time tracing, live metrics, cost tracking, and production monitoring. The observability platform for AI agents.
Total Runs
1,247
Avg Latency
342ms
Total Tokens
2.4M
Error Rate
0.3%
Active Sessions
Research Agent
Active 1m ago
RAG Pipeline
Active 2m ago
Analyst Tool
Active 3m ago
Trace Tree Explorer
Zero to Trace in seconds.
One import, one function call. Pure simplicity.
import {collector, trace }from '@orka-js/collector'; // Start the collector dashboard const { tracer } = await collector({ port: 3001 }); // Wrap your agent calls const result = await trace.wrap('agent', 'research', async () => { return agent.run('Analyze market trends'); }); // → Open http://localhost:3001 to see traces
Everything you need.
Nothing you don't.
Trace Viewer
Visualize execution trees with nested runs, latency, and costs.
Live Metrics
Monitor latency, token usage, error rates, and costs in real-time.
Remote Tracing
Send traces from production to a central collector. Debug live issues.
Session Tracking
Organize traces into sessions and export execution history.
SSE Streaming
See traces appear in real-time as your agents execute.
Zero Config
One import, one function call. Works with all OrkaJS components.
Built for the Full Stack.
ReAct Agent
Autonomous reasoning loops
RAG Pipelines
Context-aware knowledge retrieval
StateGraph
Complex workflow orchestration
Multi-Agent
Collaborative agent systems
Workflows
Sequential task automation
Tools
External API & function calling
Ready to observe your AI apps?
Install @orka-js/collector and start visualizing your LLM traces in minutes.