#System Status

#Overview

The Status page provides a read-only view of platform and project health. The page route is shared, but what it shows depends on your role:

• Client roles (Client Admin, Client Manager, Client Staff) see the platform incident status indicator, in-scope project health, active incidents, and recently resolved incidents. This is the experience described in the sections below.
• Partner roles (Partner Admin, Partner User) see the partner Status & capacity Monitor view instead. It surfaces a confirmed-client directory, source-health state, and read-only incident notices, with per-project health and live capacity shown once a client is selected. It does not show the platform incident indicator or the Active Incidents / Recent Incidents sections. See Partner Status & capacity below.

For client roles, the platform incident status is derived from active incidents:

• All Systems Operational when there are no active incidents.
• Minor Issue when only minor incidents are active.
• Service Degradation when at least one major incident is active and no critical incident is active.
• Critical Disruption when at least one critical incident is active.

When the portal cannot determine the current incident state at all, the indicator shows Status temporarily unavailable. When status evidence, recent incident history, or project-health evidence cannot be loaded or has gone stale, the indicator shows Status evidence temporarily unavailable with a short warning line below it. In-scope project health can also raise the displayed state to Service Degradation when a project health signal is degraded. An active Service Degradation or Critical Disruption is never downgraded by an evidence gap — the page keeps the more severe label and adds the warning.

#Prerequisites

• You are signed in to the portal with a supported client or partner role.
• Open the direct route /status. Client shells (Client Admin, Client Manager, Client Staff) do not include a Status sidebar item, so the direct link is the way in. Partner shells open the same route from Status & capacity in the sidebar Monitor section.

#Viewing System Status

The following sections describe the client-role experience. Partner roles see the Partner Status & capacity view instead.

1. Open the Status page at /status. Client roles reach it by direct link only.
2. Review the top status indicator. The page refreshes current incident status and project-health status through the portal about once per minute.
3. If the page shows that evidence is temporarily unavailable or stale, wait and retry before relying on the overall status. Reloading the browser tab can still return the same platform incident result for a few seconds, because the portal server briefly caches that response.

#Reviewing Project Health

1. Review the Project health section below the top status indicator.
2. Each in-scope project shows an overall state and health rows for phone path, booking, transfers, SMS, email, reports, recording, transcript, and incidents when that evidence is available.
3. Health rows can show last success, last failure, recent counts, failure rate, delivery counts, stale evidence, and source-unavailable warnings.
4. Use Refresh in the Project health section to reload project-health evidence without refreshing the whole page. If project health could not be loaded at all, the section replaces Refresh with an unavailable notice and a Retry action.
5. When a project or health row needs tenant-specific follow-up, the page can show Contact support, which opens the support form pre-filled with the project, area, and reason. Platform-wide source outages and rows reported as not provisioned do not show per-tenant support actions.
6. If no projects are in scope for your account, the section shows that no in-scope projects are available.

#Reviewing Active Incidents

1. If an incident is in progress, it appears in the Active Incidents section below Project health. The section is hidden entirely when no incident is active.
2. Each active incident shows a severity label, title, message, and created timestamp.
3. Active incident banners can also appear above portal pages. Critical banners are not dismissible; minor and major banners can be dismissed for the current browser session unless the incident is updated. The portal shows up to three active banners and links to the Status page when more are available.

#Reviewing Recent Incidents

1. Scroll to the Recent Incidents (Last 7 Days) section to see up to 20 incidents that were resolved within the last 7 days.
2. Each resolved incident shows its severity, title, message, resolved timestamp, and total duration.
3. If recent incident history cannot be loaded, the page shows an unavailable notice with a Retry action while the current status panel can still load separately.

#Partner Status & capacity

Partner Admin and Partner User roles open the same /status route but see a partner Status & capacity Monitor view rather than the client status shell. This view is read-only and scoped to the clients assigned to your partner account.

1. With no client selected, the page shows your confirmed clients as a directory. Per-client health is not asserted at this level; each client shows a prompt to select it for health detail. A source-degraded strip appears when the underlying health source reports a degraded or unavailable state. Capacity and jobs and rebuilds are gated at this level and require a selected client.
2. Provider & runtime health and Runtime jobs & rebuilds are read-only and are shown as upcoming where no partner-safe source is available yet. Jobs and rebuilds are display-only — there is no job control.
3. Partner-visible incidents appear as read-only notices. Governance and actor events are not shown here; those remain on the Audit log.
4. Select a client to see per-project health rows for that client, alongside the capacity summary (active calls, utilisation, and the latest snapshot, with a link to the full capacity page) and the provider, jobs, and incident panels.
5. If no clients are assigned to your account, the view shows that none are assigned. If the health source cannot be loaded, the view shows a recoverable unavailable state with a retry rather than an empty "no data" result.

Project-health evidence for this view comes from a single bounded read of /api/status/project-health; it does not perform a per-client routed fan-out.

#Security and Scope

The browser uses your signed-in portal session. You do not need to enter or send an API key for the status page.

Status requests go through portal BFF routes at /api/status, /api/status/recent, and /api/status/project-health. The browser calls the portal only; the portal server forwards your session bearer token to aidial_api. The page route is protected by portal session middleware, and the API routes enforce their own session and RBAC checks.

Platform incident status is a shared portal surface, while Project health is scoped to projects available to your signed-in account. Tenant and project scope is resolved server-side, and unsupported roles or out-of-scope requests are blocked without exposing unrelated operational detail.

#Common Issues

• The status page shows everything as operational but I am experiencing problems. Localised issues may not always appear on the status page immediately. If the problem persists, contact your partner representative for support.
• I want to be notified when an incident occurs. Incident notifications are not currently managed from Notification Preferences. That settings page currently covers per-project post-call transcript email and SMS summary delivery; monthly summary delivery is managed separately on Reports.
• An incident was resolved but I am still affected. Some issues may require you to refresh your browser or sign out and sign back in after resolution.
• Status is not visible in the sidebar. The client shells do not include a Status sidebar item; open /status directly. Partner shells show it as Status & capacity under Monitor. Navigation visibility is not the security boundary; the protected page and API routes still enforce session authorisation server-side, and an unauthorised role receives a not-found response rather than a listing of what exists.

#Related Pages

• Getting Started
• Troubleshooting