Console dashboard — bRRAIn Docs

The Console dashboard — what every card shows, how to read the health indicators, and how to drill in.

Console dashboard

The dashboard is the first page you see when you open the Console. It gives you a 30-second read on your organization's state.

At-a-glance cards

Five cards along the top:

  1. Pod status — green / amber / red pip plus version, region, and uptime. Click for the full pod detail page.
  2. Members — total count and any pending invitations. Click to go to the Members page.
  3. Storage — Vault usage and Document Portal usage as a stacked bar. Click for storage breakdown.
  4. This month's spend — accruing usage charges for the current billing cycle. Click for billing detail.
  5. Open alerts — count of unresolved notifications across pod, billing, and integrations. Zero is the normal state.

Activity feed

Beneath the cards, the activity feed shows the last 50 noteworthy events:

  • New members joining or leaving.
  • Marketplace extension installs, updates, uninstalls.
  • Integration changes (OAuth credentials added, webhooks configured).
  • Pod lifecycle events (restart, upgrade, region change, scale).
  • Policy changes (retention rules, role assignments, quota changes).

Each entry shows the actor, the action, the affected resource, and the timestamp. Click any entry for the underlying audit-log record.

Compute · Over Time chart

A line chart of brain-pod compute usage with three series:

  • CPU % — average across the last sample window.
  • GPU % — present when your pod has a GPU; pulls from nvidia-smi-equivalent telemetry.
  • Inflight requests — concurrent requests being served.

Range selector: 1h / 24h / 7d / 30d. Hover for exact values.

This chart is your first stop if members are reporting slow responses or if you suspect your pod is undersized.

Recent searches and queries

A small panel showing the most-recent semantic searches against the Vault, with the requester (anonymized to role for privacy) and result count. Useful for spotting query patterns that might benefit from new ontology entries or a domain-adapter swap.

Open approvals

If you have approval gates configured (see Roles & permissions), this panel shows the queue. Each row links to the underlying action with Approve and Deny buttons. Approvals are also surfaced in your bell notifications and via email.

Quick actions

A vertical strip on the right with the most-used operator actions:

  • + Invite member
  • + Install extension (deep-links to the Marketplace)
  • Restart pod
  • Open audit log
  • Take pod backup

Each action carries you to the relevant page with the right context preloaded.

How fresh is the data

  • Pod-status cards refresh every 15 seconds.
  • The activity feed is push-updated via server-sent events; you'll see entries appear without refreshing the page.
  • The compute chart polls once per minute on the 1h range and proportionally less often on longer ranges.
  • Storage and spend cards refresh every 5 minutes.

If a card shows a stale-data warning, the underlying source is unreachable; the displayed number is the last known good value, and the timestamp is shown.

Customizing the dashboard

Most cards can be reordered or hidden via Settings → Dashboard layout. Layout preferences are per-user — your Architect colleague can show different cards than you.

Mobile

The dashboard is responsive. Cards stack to a single column on phones. The activity feed and compute chart remain available; the quick actions become a floating action button.

Where to next