Stop slipping deals: triage Early Warning

“Spot slipping deals before forecast surprises.” Work the queue, claim what is yours, and resolve every case with an auditable decision.

What you’ll achieve

An empty-enough queue: every warning claimed, acted on, snoozed with intent, or dismissed with a reason — none of them scrolling silently into next month.

Why this matters

A warning nobody owns is a loss with extra steps. The entire value of early warning collapses if cases sit unclaimed — the discipline of triage is what turns detection into saved revenue.

The steps

1
Open Early Warning and read the triage bar
The workspace opens with the numbers that matter: SLA breached, Unowned, Critical risk, My queue, Exposure. That bar is your triage order — breaches and unowned cases first.
2
Pick your view
Spotlight shows the top-priority case with the queue below; Queue is the full refined list; Patterns shows trends and clustering; Split puts list and detail side by side. Saved chips — “My queue”, “SLA breached”, “Unowned”, “High exposure” — get you to the right slice in one click.
3
Read a case: what happened, why it matters
Every case is written in plain language: “What happened”, “Why it matters”, “Outcome at risk”, “Deal value exposed”, and a recommended action. “Proof behind this” shows the evidence the warning is built on.
4
Claim it or route it
Click “Claim” to own a case, or assign it out. Unowned is the worst state a warning can be in — the triage bar counts them for a reason.
5
Act, then close the loop
Take the recommended action (or your better one), then record the outcome: “Acknowledge” while working it, “Escalate” to pull in help, “Resolve” when handled.
- “Snooze” (3, 7, or 14 days) hides a case and brings it back — use it for “real, but not this week”.
- “Dismiss” is final and requires a note. Every action is auditable: your team sees your latest decision, not raw model output.

You’re done when

Nothing in the queue is unowned.
Every case you touched left with an action, a snooze, or a reasoned dismissal.

Next guide

Start your day from the Pending inbox

Do it in your own workspace

Use this guide in an approved private-beta workspace after connecting a supported source; availability and setup depend on your data and configuration.

Request beta access Run a pipeline audit

Private beta access is invite-only and requires approval.

The steps

Open Early Warning and read the triage bar

The workspace opens with the numbers that matter: SLA breached, Unowned, Critical risk, My queue, Exposure. That bar is your triage order — breaches and unowned cases first.

Pick your view

Spotlight shows the top-priority case with the queue below; Queue is the full refined list; Patterns shows trends and clustering; Split puts list and detail side by side. Saved chips — “My queue”, “SLA breached”, “Unowned”, “High exposure” — get you to the right slice in one click.

Read a case: what happened, why it matters

Every case is written in plain language: “What happened”, “Why it matters”, “Outcome at risk”, “Deal value exposed”, and a recommended action. “Proof behind this” shows the evidence the warning is built on.

Claim it or route it

Click “Claim” to own a case, or assign it out. Unowned is the worst state a warning can be in — the triage bar counts them for a reason.

Act, then close the loop

Take the recommended action (or your better one), then record the outcome: “Acknowledge” while working it, “Escalate” to pull in help, “Resolve” when handled.

“Snooze” (3, 7, or 14 days) hides a case and brings it back — use it for “real, but not this week”.
“Dismiss” is final and requires a note. Every action is auditable: your team sees your latest decision, not raw model output.