Using the dashboard

Complete reference for app.amlexia.com. Requires Clerk login; uses your org’s projects.

Header controls

Control	What it does
Org switcher	Change organization context
Project switcher	Select which project’s data you view
Time range	1h, 24h, 7d, 30d — applies to most analytics pages
Environment	All, production, staging, development — filters by event environment field
Compare	Toggle side-by-side comparison of two environments
Saved views	Save current range + env + compare; load or delete presets

TIP

Set AMLEXIA_ENVIRONMENT in every SDK deployment so filters work.

Monitor

Overview (/dashboard)

• Stat cards: request volume, error rate, avg latency, cost
• Cost breakdown: reported vs estimated LLM cost
• Tabs: Overview charts, Latency table, Errors table, Cost table
• CSV export on tabular data
• Empty state links to Settings for SDK key if no data

When Compare is on, an environment comparison panel replaces single-env charts.

Live (/dashboard/live)

• Server-Sent Events stream of incoming events
• Best for first integration and deploy verification
• See Live stream

Traces (/dashboard/traces)

• List traces for selected range/environment
• Open trace detail → spans, linked events
• Requires SDK tracing or OTEL — Tracing guide

Intelligence

Providers (/dashboard/providers)

• Card per detected provider: uptime, latency, retries, 24h cost
• AI Models table: per-model requests, tokens, P95, cost
• Respects global time range and environment filter

Sessions (/dashboard/sessions)

• User/session groupings when session_id / user_id sent on events
• Pass x-session-id / x-user-id headers in middleware

Insights (/dashboard/insights)

• Insight cards with severity and recommended actions
• Links to Operations, Alerts, Traces
• AI summary card — narrative health (12h window)
• Dismiss resolved insights

See Insights & anomalies

Operations (/dashboard/operations)

• SLO availability vs target
• P95 latency
• 30-day LLM spend forecast
• Provider health overlay
• Incident timeline

See Operations

Platform

Infrastructure (/dashboard/infrastructure)

• Infra-oriented metrics when provider is aws, supabase, firebase, etc.

Captures (/dashboard/captures)

• Stored request/response captures (when enabled in project settings)
• Replay for debugging

Manage

Projects (/dashboard/projects)

• List/create projects
• Switch active project

Alerts (/dashboard/alerts)

• Create/edit/delete alert rules
• Toggle active/inactive
• Alert history with trigger details JSON
• Channels: email, Slack webhook, generic HTTP webhook

See Alerts

Settings (/dashboard/settings)

Section	Contents
SDK key	Copy, regenerate (am_...)
Project	Name, retention, capture toggles
Team	Members, roles (owner/admin/member)

Regenerating a key immediately invalidates the old key.

Sidebar usage meter

Shows events this month / plan limit (e.g. 3,240 / 10,000 on Startup). Over limit → ingest 402. See Plans & limits.

Onboarding & product tour

New users see onboarding until the first event. Product tour replays from the header menu.

Mobile

Hamburger menu exposes navigation on small screens.

API equivalents

Every view is backed by Dashboard API endpoints if you build custom tooling (still requires Clerk JWT, not SDK key).

Related

• Complete setup
• Dashboard overview