Digital Twin

A Digital Twin is a causal model of your business or system. Unlike traditional predictive models that only learn correlations, a Digital Twin understands cause and effect—enabling simulation, optimization, and counterfactual reasoning.

---

Building a Digital Twin

Creating a Digital Twin

Start here. Learn how to create a Digital Twin from a Data View, choose between static and temporal models, and run causal discovery.

---

Understanding the Model

Causal Graph

The causal graph shows discovered cause-and-effect relationships. Learn to read, interpret, and refine the graph with domain knowledge.

---

Using the Model

Simulations

Run what-if scenarios, optimize decisions, find root causes, and predict outcomes. This is where causal understanding becomes actionable.

---

Comparing Models

Model Comparison

Compare different versions of your Digital Twin or compare models across twins. Understand exactly what changed in structure and parameters.

---

The Digital Twin Interface

Once you've created a Digital Twin, you interact with it through several tabs:

Digital Twin Tabs

• Home Tab – Overview, version selection, quick actions

• Config Tab – Data view selection, field configuration, constraints

• Relationships Tab – View and edit causal relationships

• Path Analysis Tab – Sankey diagrams showing causal flow

• Evaluation Tab – Model quality metrics

• Simulation Tab – Run all simulation types

• Comparison Tab – Compare model versions

---

Workflow

A typical Digital Twin workflow:

1. Create – Select a Data View and run causal discovery

2. Review – Examine the discovered causal graph

3. Refine – Add domain knowledge, resolve ambiguous edges

4. Evaluate – Check model quality metrics

5. Simulate – Test interventions, optimize, explain outcomes

6. Iterate – Improve the model based on what you learn