Use Google, GitHub, or magic link. Successful OAuth login returns to the homepage unless a next parameter is supplied.
Run UmamiEdge as a global distributed AI compute pilot.
This guide explains the product workflow for founders, operators, site owners, enterprise customers, and internal teams using UmamiEdge v8.
Operator shortcuts.
From login to pilot launch.
Go to onboarding, create a tenant organization, and assign the first owner/admin account.
Register a commercial site such as a colocation room, telecom edge room, university, bank campus, hospital, solar mini-grid, or enterprise facility.
Create the first GPU node, generate a node API key, and store the secret securely because only the prefix/hash is retained.
Use Mission Control, Ops Center, Nodes, Telemetry, Incidents, SLA, and Billing Control to track health and commercial readiness.
Use the Gateway and customer API-key pages to prepare OpenAI-compatible integrations and regional routing policies.
Dashboard concepts.
Node health
Online, degraded, offline, and retired node states show operational availability and maintenance risk.
Telemetry
GPU utilization, temperature, latency, power draw, and heartbeat events indicate whether the node is safe for inference workloads.
Routing
The router selects nodes based on health, latency, available power, residency rules, runtime readiness, and tenant quota.
Incidents
Warnings and critical telemetry should create or update incident records for operator triage and audit visibility.
Billing
Usage events, committed capacity, node revenue, and site-owner payouts define pilot economics.
Trust evidence
Compliance packs summarize node authentication, RLS, data residency, SLA policy, and incident response posture.
Who uses what.
Owner
Controls organization settings, billing, team access, customer keys, and strategic pilot configuration.
Admin
Manages sites, nodes, runtimes, customer keys, compliance evidence, and billing-control workflows.
Operator
Monitors telemetry, handles incidents, schedules maintenance, and runs failover/SLA checks.
Viewer
Reads dashboards, reports, guides, and executive status without changing production settings.
Site owner
Tracks site earnings, power consumption, uptime, maintenance windows, and contract status.
Customer
Uses OpenAI-compatible APIs, quotas, model endpoints, and SLA reports.