Quickstart
First trace in 5 minutes.
Three steps from install to live agent data in your RouteIQ dashboard. No infra changes required.
Python 3.9+Any OTel CollectorAny agent framework
1Install
# Core SDK pip install routeiq # With auto-instrumentation (LangChain, OpenAI, Anthropic, CrewAI) pip install "routeiq[traceloop]"
2Configureconfigure.py
import routeiq routeiq.configure( agent_id="support-agent-prod", api_key="riq_live_..." # from routeiq.dev endpoint="http://localhost:4318", # your OTel Collector enable_traceloop=True, )
3Instrumentagent.py
@routeiq.instrument async def run_agent(query: str) -> str: # Your agent code unchanged — LLM + tool spans auto-captured result = await your_agent.ainvoke(query) return result
That's it. Here's what's flowing:
- agent_start and agent_end spans on every run
- LLM and tool spans auto-captured via Traceloop
- Data in ClickHouse and your RouteIQ dashboard
- SLO monitoring active from the first trace
Framework Integrations
Works with your stack.
One OpenTelemetry endpoint. Every framework, every cloud, every model.
LangGraphOpenAI Agents SDKCrewAILangChainLlamaIndexCustom runtimes
Next steps
Where to go from here.
SDK Reference
Full API docs for routeiq.configure(), @routeiq.instrument, sessions, tools, and escalations.
Schema Reference
ClickHouse table layouts, span types, log attributes, and metric names.
Setting up SLOs
Define task success, latency, cost, and policy thresholds for your agent.
OTel Collector Setup
Configure the pipeline for local, VPC, or cloud-managed environments.
Stuck? We'll help.
RouteIQ is in private beta. Reach out and we'll help you instrument your specific agent stack.