SLO dashboard and error-budget burn alerts

The SLO dashboard tracks the error-budget burn rate for your AI agents against an availability target, so you can see when the budget is burning faster than expected — before you miss it for the month.

Core concepts

Term	Definition
SLI	Service Level Indicator — the metric being measured (e.g. success rate)
SLO	Service Level Objective — the target (e.g. 99% success over 30 days)
Error budget	100% − SLO target. At 99% SLO, 1% of requests can fail per month.
Burn rate	How fast the error budget is being consumed relative to the budget window.

A burn rate of 1.0 means you are exactly on track to use the full budget by the end of the window. A burn rate of 14.4 means you will exhaust the monthly budget in about 50 hours (a 30-day window) if the current rate continues.

Alerting

Burn-rate alerting is live. Create a rule in Settings → Alerts, set a threshold, and breaches are delivered to a webhook you control — Slack, Discord, or any HTTPS endpoint. Alongside burn rate you can alert on error rate, p95 latency, gateway overhead p99, spend, and quota used. Alerting is available on paid plans. Per-SLO configuration (custom SLIs, targets, windows) is not built — the dashboard tracks the burn rate against a default availability target, and the target is per plan, not per SLO.

Dashboard

The /slo page in the dashboard shows:

Error budget remaining — percentage and absolute count against the availability target.
Burn rate — how fast the error budget is being consumed relative to the window.
Per (provider, model) latency — p50 / p95 / p99 and error rate from the hourly rollups.
Top failing requests — the traces that contributed most to budget burn, linked directly to the trace viewer.

API

curl "https://gateway.tracelane.dev/v1/slo" -H "Authorization: Bearer tlane_..."

Returns hourly SLO rollups per (provider, model):

[
  { "bucket_hour": "2026-07-15T14:00:00Z", "provider": "anthropic",
    "model": "claude-sonnet-4-6", "p50_ms": 812.0, "p95_ms": 2140.0,
    "p99_ms": 5300.0, "requests": 1284, "errors": 9,
    "error_rate_pct": 0.70, "total_input_tokens": 0, "total_output_tokens": 41233 }
]

Burn rate is derived client-side against your plan’s availability target (99% Team, 99.9% default, 99.95% Enterprise) (apps/web/app/slo/budget.ts); there is no per-SLO objective field.

Get started

SDKs

Integrations

Concepts

Features

Tools

Audit & compliance

Operations

Migrations

Comparisons

SLO dashboard and error-budget burn alerts

Core concepts

Alerting

Dashboard

API

​Core concepts

​Alerting

​Dashboard

​API

​Related

Core concepts

Alerting

Dashboard

API

Related