Back to blog
Pulse

How Pulse turns test runs into release confidence

Pass rate is not confidence. Pulse blends per-area execution snapshots with live Blueprint coverage, surfaces Likelihood × Impact risk, and keeps KPIs, heatmap, and trends on one lens.

Pulseriskhealthcoveragerelease

Nimra Naseem

Senior QA Engineer

3 min read

Executives ask a simple question before ship: how confident are we? Engineering answers with dashboards of pass rates. Those numbers feel precise and often mislead.

Pass rate collapses many stories into one: a flaky area and a rock-solid area can look identical if both “mostly pass.”

Two sources, one dashboard

Pulse is a continuous health dashboard for your product, scored per Blueprint area. It blends two sources that update on different cadences:

SourceWhat it measuresWhen it updates
ExecutionHow linked tests actually ran in orchestrationsWhen a run finishes and Pulse stores a snapshot
CoverageSpec↔area and spec↔test link health from your BlueprintLive when you change links — no new run required

The dashboard opens after at least one finished orchestration produces an execution snapshot. Coverage then layers into the KPI strip and trends; the area heatmap colours from execution health so the map stays tied to what actually ran.

The lens keeps every panel honest

The lens is the saved view of the dashboard. It chooses:

  • Which health signals feed the blend (test execution, spec–area, spec–test)
  • Environment, range (7d / 30d / 90d / custom), and tags
  • Which runs on the timeline you are comparing

KPI strip, heatmap, and trend chart stay aligned to that lens — so you are never comparing apples to oranges across panels.

States you can defend in a go/no-go

Each area lands in Healthy, Warning, Critical, Unknown, or Ignored (archived areas stay out of the overall score). Colour is never the only cue: labels and icons keep the board usable under pressure.

Execution snapshots also carry a Likelihood × Impact risk read (1–25) for the highest-area-risk KPI — useful when pass rate looks fine but one critical subtree is concentrating danger.

Break the trend chart down by:

  • Overall project health
  • Orchestration (which workflow is hurting the score)
  • Area (where risk concentrates)
  • By health score when coverage signals are in the lens — including days with no execution snapshot

Drill a heatmap tile and the trend re-scopes to that subtree. Clear the selection to return to project-wide view.

Wiring Pulse into the release ritual

  1. Before code freeze: close Gaps on high-criticality Blueprint nodes.
  2. During the release candidate: watch orchestration’s live area heatmap.
  3. At go/no-go: open Pulse with a lens that matches the release environment and tags — not only the CI badge.
  4. After ship: keep the snapshot history so the next cycle has a baseline.

What changes culturally

Teams stop arguing about “enough tests” in the abstract. They argue about which areas need coverage or healthier runs before the next train. That is a healthier conversation — and it scales with AI-authored change.

Pulse does not replace judgment. It gives judgment a shared, evidence-backed language.