MODULES

Metric–State Engine

Decide when data becomes actionable

This module determines when operational signals cross the line into a real business issue.

The Metric–State Engine is the entry point of decision-making in Runbook Metrics.

It continuously evaluates structured operational data against explicit rules to determine when attention is required

What it does

Operational data such as sales performance, inventory movement, margin shifts, returns, or supply signals is continuously evaluated against configurable rules.

Instead of reacting to raw fluctuations, the engine assesses patterns, thresholds, trends, and persistence over time — translating data behaviour into clear operational states such as Normal, Warning, or Action Required.

Why it matters

Not every change deserves attention.

By distinguishing noise from meaningful signals, the engine ensures only issues that truly matter surface.

This results in fewer false alarms, clearer prioritisation, and faster alignment on what needs action.

Output

• Clear operational states reflecting business impact
• Risk levels used to drive downstream actions

• 
  

Impact

Managers and operations no longer ask “Is this normal?” — the system already answers that.

Runbook Engine

Define what must be done — not debate how

This module determines what action should be taken once intervention is required.

The Runbook Engine is where decisions turn into execution.

It transforms operating procedures into executable, accountable actions.

What it does

Rather than treating SOPs as static documentation, the Runbook Engine treats them as executable configuration. 

Instead of treating SOPs as static documents, the Runbook Engine treats them as structured configurations.

Each actionable state is mapped to a predefined runbook that specifies:

• Who is responsible

• What steps must be taken

• Expected response timeframe

Whether it’s replenishment, promotion review, supplier follow-up, or investigation, responses are explicit and consistent.

Why it matters

Without clear responses, signals lead to discussion instead of action.

This layer removes ambiguity and ensures the right action happens, even when teams or shifts change.

Output

• A concrete action tied to the current state
• Clear ownership and response expectations
• 
  

Impact

Execution no longer depends on experience or tribal knowledge. The right action is clear, every time.

AI Insight Layer

Explain what happened — and what deserves attention

This module determines how quickly a situation can be understood.

The AI Insight Layer exists to explain outcomes clearly and consistently, without replacing rules or human judgment.

It explains why something changed, what matters, and what stands out — both in response to events and through regular operational summaries.

What it does

When data enters a warning or action-required state, the AI analyses historical context and current behaviour to explain:

• Why the situation occurred

• What changed materially

• Which drivers and risks are most relevant

In addition to event-based explanations, the same layer generates periodic digest summaries (daily or weekly), highlighting emerging patterns, notable changes, and top risks — even when no single alert is triggered.

Why it matters

Manual interpretation is slow and inconsistent.

By standardising explanation, teams start from the same understanding and spend less time analysing — and more time acting.

Output

• Human-readable explanations of why a trigger occurred
• Clear summaries of key drivers, risks, and changes
• 
  

Impact

Reduced time spent on analysing reports and more time acting.

Action & Alert Delivery

Ensure decisions reach the right owner

This module determines whether decisions actually lead to action.

A decision that is not delivered in time has limited operational value.

What it does

Instead of generic alerts, this layer delivers structured, action-oriented messages that clearly state:

• What triggered

• Why it matters

• Which action applies

• Who owns it

• When a response is expected

Messages are delivered in formats ready for execution, with links to supporting context.

Why it matters

Operators and managers shouldn’t need to hunt for information before acting.

By delivering decisions directly, this layer shortens response time and reduces missed actions.

Output

• Action-driven notifications or briefings
• Clear ownership, priority, and response expectations
• 
  

Impact

Issues are handled faster, and fewer things fall through the cracks.

Visualization Layer

Provide clear visibility into the data behind every decision

This module gives teams a clear, shared view of the data that underpins every metric, state change, and recommended action.

In Runbook Metrics, dashboards serve as the evidence layer that supports operational decisions. 

They are designed to make data understandable, reviewable, and traceable — enabling confidence in both real-time response and post-action review.

What it does

Curated visualisations show how conditions evolved over time, when states changed, and how actions unfolded — using the same data that drives rules and AI explanations.

Teams can review trends, validate outcomes, and trace actions back to source data when needed. 

Why it matters

This layer enables confidence without slowing execution.

Teams can review, learn, and audit decisions without manual reporting or constant monitoring.

Output

• Clear visual context for every decision
• Historical traceability across metrics, states, and actions

Impact

It allows managers and operators to see what changed, where it happened, and why — without pulling reports or chasing evidence.