The whole execution path
Follow retrieval, model calls, tool invocations, and policy checks as one connected timeline.
Connected observability
See the signal.
Follow the story.
Every model request is part of a bigger execution. Bring the trace, the context, and the cost into the same view.
Total requests1.28M+18.6% vs. previous period
Success rate99.98%+0.04% vs. previous period
P50 latency142ms-12 ms vs. previous period
Estimated cost$84.62-8.2% vs. previous period
Request volume
Last 24 hoursSuccessful requests1.28M total
Model routing
Atlas / fastPrimary route
68%Solace / reasonComplex tasks
24%Ember / compactFallback route
8%Active deployments 3
All healthy| Service | Region | Status | Requests | Latency |
|---|---|---|---|---|
| support-resolver | us-east-1 | Healthy | 842.1k | 142 ms |
| document-pipeline | eu-west-1 | Healthy | 318.6k | 208 ms |
| research-orchestrator | ap-south-1 | Healthy | 124.9k | 186 ms |
Support resolution flowv2.4.1 / Published
TRIGGERNew support requestwebhook / support.inboundListening
AGENTResolve with contextAtlas / fast + knowledge base98.4% resolved
GUARDRAILValidate responsepolicy / customer-safePolicy passed
Last run completed3 steps · 1.42s · $0.0021
EXECUTION TIMELINE
Context retrieval210 ms
Agent reasoning980 ms
Policy check230 ms
Trace completedsupport-resolvertr_8f2c91 · 142 ms
POST /v1/agents/runtr_8f2c91
142 ms200↳retrieve.contextsp_72a801
38 ms200↳model.generatesp_6b3f42
89 ms200↳policy.validatesp_4d8e12
15 ms200ATTRIBUTES
- environment
- production
- region
- us-east-1
- model
- atlas-fast-2
- policy
- customer-safe
EVENTS
Request accepted
Context retrieved · 4 documents
Model response received
Policy passed · response sent
Estimated spend / last 24h$84.628.2% below previous period
Cost by workload
USDsupport-resolver$48.12
document-pipeline$24.36
research-orchestrator$12.14
Budget policy active. Alert at 80% of your $3,000 monthly cap.
All systems operationalSample workspace · illustrative data
Illustrative execution trace. No customer or production information is collected.
01Less searching. More understanding.
A trace that answers
the next question.
Latency in the right place
Separate time spent retrieving context from time spent reasoning. Fix the slow step, not the wrong service.
Cost with attribution
Connect spend to a workflow, a model route, or a workspace. Make trade-offs with a useful unit of work.
02Context at the point of investigation
From an alert
to an explanation.
A spike should lead to a question you can answer. Keep events connected to the versions, policies, and workspaces that produced them.
- Trace IDs connect each agent handoff.
- Policy decisions appear next to tool activity.
- Model route changes remain visible in context.
FROM EVENT TO UNDERSTANDING01 / 03
Agent execution
Model request
Tool invocation
A connected trace
Latency signals
Cost attribution
Actionable alerts
TRACEABLE BY DEFAULTCONTROLPLANE
JSON / illustrative event
{
"event": "agent.run.completed",
"traceId": "tr_8f2c91",
"workflow": "support-resolver",
"version": "2.4.1",
"durationMs": 142,
"costUsd": 0.0021,
"policy": "customer-safe",
"outcome": "resolved",
"redaction": "applied"
}03Signals that travel
Your telemetry.
Still yours.
Keep a clean event contract between the operating layer and the tools your team uses to investigate it.
Choose what belongs in a trace. Redact sensitive values before exporting events, and separate useful attributes from raw customer input.
Read the event contractFrom first request to production
Keep the context.
Lose the guesswork.
Make the next investigation start with a connected view.