AI (artificial intelligence) predicts and generates; BI (business intelligence) reports what happened; CI (customer intelligence) explains who customers are and what they will do; CX intelligence shows how the experience feels and what to fix first; SI (scenario intelligence) tests what-ifs; OI (operational intelligence) shows what is happening right now; and DI (decision intelligence) decides what to do, routes it to the right authority, executes it and measures the outcome. They stack: data feeds BI and OI, AI and SI add prediction and simulation, and DI turns all of it into accountable decisions.
Every kind of intelligence, side by side
| Name | The question it answers | What it produces | |
|---|---|---|---|
| AI | Artificial intelligence | What is likely, and what does this mean? | Predictions, classifications, generated text and images |
| BI | Business intelligence | What happened, and where? | Reports, dashboards, KPIs |
| CI | Customer intelligence | Who are our customers and what will they do? | Segments, propensity and churn scores, 360° profiles |
| CX | Customer-experience intelligence | How does it feel to be our customer — and what should we fix first? | Journey analytics, sentiment, next-best-action |
| DI | Decision intelligence | What should we do, who decides, and did it work? | Decisions executed under policy, with outcomes measured |
| SI | Scenario intelligence | What happens if…? | Simulations, what-if comparisons, stress tests |
| OI | Operational intelligence | What is happening right now? | Real-time monitoring and alerts on live operations |
| MI | Market & competitive intelligence | What are markets and competitors doing? | Market sizing, competitor moves, pricing signals |
| SCI | Supply-chain intelligence | Where will supply break, and what then? | Supplier risk, lead-time forecasts, network visibility |
| Physical AI | Physical AI | How should a machine act in the real world? | Robot and vehicle actions inside safety envelopes |
| Agentic AI | Agentic AI | Can software carry out a multi-step task itself? | Agents that plan and act across tools |
| Edge intelligence | Edge intelligence | Can the decision be made where the data is? | Models and rules running on devices and sites |
How they fit together
↺ Outcomes flow back into the data — the loop that makes every layer smarter.
None of these replaces the others. BI without DI is a report nobody acts on. AI without DI is a prediction with no owner. DI without BI and AI has nothing to reason with. The value is in the stack — and in closing the loop from action back to data.
AI — Artificial intelligence
The question it answers: What is likely, and what does this mean?
What it produces: Predictions, classifications, generated text and images.
Who uses it: Data scientists, every application.
Example: Forecasting demand; reading a document; spotting a defect in an image.
BI — Business intelligence
The question it answers: What happened, and where?
What it produces: Reports, dashboards, KPIs.
Who uses it: Analysts, managers.
Example: Monthly sales by region; OEE by line.
CI — Customer intelligence
The question it answers: Who are our customers and what will they do?
What it produces: Segments, propensity and churn scores, 360° profiles.
Who uses it: Marketing, sales, product.
Example: Which accounts are likely to churn next quarter.
CX — Customer-experience intelligence
The question it answers: How does it feel to be our customer — and what should we fix first?
What it produces: Journey analytics, sentiment, next-best-action.
Who uses it: CX, support and service leaders.
Example: Routing a frustrated caller to a senior agent with the fix ready.
DI — Decision intelligence
The question it answers: What should we do, who decides, and did it work?
What it produces: Decisions executed under policy, with outcomes measured.
Who uses it: Operators, managers, executives — and machines.
Example: Swap a compressor bearing in this week’s slowdown; reliability engineer approves.
SI — Scenario intelligence
The question it answers: What happens if…?
What it produces: Simulations, what-if comparisons, stress tests.
Who uses it: Planners, finance, strategy.
Example: Margin impact of losing a supplier for three weeks.
OI — Operational intelligence
The question it answers: What is happening right now?
What it produces: Real-time monitoring and alerts on live operations.
Who uses it: Control rooms, operations.
Example: Live line status and alarms across three plants.
MI — Market & competitive intelligence
The question it answers: What are markets and competitors doing?
What it produces: Market sizing, competitor moves, pricing signals.
Who uses it: Strategy, product marketing.
Example: A competitor’s price cut in two regions.
SCI — Supply-chain intelligence
The question it answers: Where will supply break, and what then?
What it produces: Supplier risk, lead-time forecasts, network visibility.
Who uses it: Procurement, planning, logistics.
Example: A port delay that will hit orders in 12 days.
Physical AI
The question it answers: How should a machine act in the real world?
What it produces: Robot and vehicle actions inside safety envelopes.
Who uses it: Robotics, automation and operations teams.
Example: An AMR slowing and rerouting to service.
Agentic AI
The question it answers: Can software carry out a multi-step task itself?
What it produces: Agents that plan and act across tools.
Who uses it: Operations, IT, service.
Example: An agent that reconciles an invoice mismatch end to end.
Edge intelligence
The question it answers: Can the decision be made where the data is?
What it produces: Models and rules running on devices and sites.
Who uses it: OT, field and device teams.
Example: A camera rejecting a part in 20 ms without the cloud.
Which one do you need?
| If your problem is… | Start with | Then add |
|---|---|---|
| “We don’t know what happened last quarter.” | BI | OI for live views |
| “We see problems too late.” | OI + AI (prediction) | DI to act on them |
| “We have forecasts but nothing changes.” | DI | SI to compare options |
| “Customers leave and we don’t know why.” | CI + CX | DI for next-best-action |
| “Our robots and machines need rules for acting alone.” | Physical AI + DI | Edge intelligence |
| “We want agents to do work end to end.” | Agentic AI + DI governance | Decision observability |
- BI looks back, AI looks ahead, DI decides and acts.
- CX and CI are about customers; DI turns their insight into action.
- The intelligences stack — value comes from closing the loop.