Automation Constellation Demo Industries AG · fictional Maturity phase 1

An organisation's automation level at a glance · 6 value streams · 67 processes · automation, AI suitability, opportunity and risk class

Colour by

Processes

–

in 6 value streams

Avg automation

–

volume-weighted –

Top-10 opportunity

–

share of total opportunity

Phase 1 · Reactive

–

manual, people react

Phase 2 · Recommendation

–

system proposes, people decide

Phase 3 · Agentic

–

executes within guardrails

Phase 4 · Autonomous

–

people monitor only

High-risk processes

–

human oversight required

Constellation map drag to pan · wheel / pinch to zoom · click a star

Stream Min opportunity 0 Risk class

zoom 1.0×

How it works

Demo Industries AG is a fictional mid-sized manufacturer (about 2,000 employees). Its six value streams and 67 sub-processes were invented for this demo; every number is synthetic and deterministic (seeded), so a reload shows the same picture. Nothing here describes a real company.

What the map shows

The core is the organisation. The inner ring holds the six value streams; each stream owns an angular sector into which its processes are fanned as stars. Star brightness encodes the automation score (dim = manual, bright = automated), star size the monthly volume, a blue ring marks the ten processes with the highest opportunity score, and a red tick marks a high-risk process. Names of the top-10 appear from 1.4× zoom, all names from 1.9×. Energy pulses travel from the core through the streams to the processes to show the direction of flow; they carry no data. Zoom in to see process names.

Scoring

Automation score (0–100): share of the process that runs without manual re-keying, assessed per process. Input data in a real assessment.
AI suitability (0–100) is rule-based from three observable traits: 0.40 × repetitive + 0.35 × digital input + 0.25 × (1 − judgement). Repetitive, digitally available, low-judgement work scores high; creative or discretionary work scores low.
Opportunity (0–100) = 100 × V^0.3 × G^0.4 × S^0.5, where V is the log-scaled volume, G the automation gap (1 − automation / 100) and S the AI suitability. High-volume, poorly automated, well-suited processes rise to the top. The top-10 share tells you how concentrated the potential is.
Risk class (low / limited / high) follows a governance view in the spirit of the EU AI Act: processes that decide about people (hiring, performance, creditworthiness), product safety or regulatory filings are high-risk. High-risk processes require human oversight and documented decision logic regardless of their technical potential.
Maturity phase placement applies the four phases of the AI maturity model to a single process: automation < 40 → 1 Reactive; 40–69 → 2 Recommendation; 70–89 → 3 Agentic; ≥ 90 with low risk → 4 Autonomous. High-risk processes are capped at phase 2 — a person must take the decision. The organisation's phase is the phase of the median process.

Feeding real data

In a real assessment the same picture is built from a CSV export of the process inventory: one row per process with stream, process, owner, volume_per_month, automation_0_100, repetitive_0_1, digital_input_0_1, judgement_0_1, risk_class, tools. The suitability, opportunity and phase columns are derived — nobody has to guess them. Thirty to ninety minutes with each stream owner is usually enough to fill the inventory; the map then becomes the agenda of the automation roadmap. This demo does not implement an upload.

Privacy: this page stores nothing, sets no cookies and sends no data; the whole computation runs in your browser. The company, processes, owners and figures are fictional.