Live Twins: A First Look

What live twins are, when to use them, and how they sit on top of your existing model. Read this first before creating a twin or entering data.

A live twin turns your model into a computational knowledge graph — the same node types and relationships you've already defined, instantiated with data that changes over time. Your model stops being a static picture and starts telling you where you've been, where you are, and where you're heading.

The starting point can be as simple as a spreadsheet of KPIs; from there, you can grow into a full agent-based simulation without changing the underlying model. The tool scales with your ambition.

When to use it

Track KPIs over time — fill in last year's actuals and next year's targets, and you have a planning spreadsheet that lives inside your model. The same setup works for ten KPIs or a thousand.
Project a scenario forward — write simple formulas and let your agents evolve automatically. Full Rhai scripting is available when you need it, so there's no upper limit to the simulations you can build.
Compare alternatives — run multiple scenarios in the same twin ("Baseline" vs "Aggressive Hiring") and put the results next to each other.
Play it through — walk a simulation forward one round at a time and decide in each round, like a game. The Beer Game template is a ready-made example (see Playing Through a Simulation Step by Step).

Two ways to fill in values

Every property on every agent has a time series — a value at each timestep. You can fill it in two ways, and you can mix them freely:

By hand in the time series grid — no formulas, no scripting. Best for historical data and planning targets.
With a formula that computes the value automatically at each step. Formulas range from one-line spreadsheet-style expressions to full Rhai scripts that read connected agents and traverse the graph — you decide how far to push it.

A common starting point is: type in last year's actuals, write a formula for next year's projection. As real data arrives later, you can compare it against what the formula predicted.

The shape of a twin

Open the Model Navigator and look under the Twins section. You'll see this structure:

Twin (e.g., "HR Operations")
  ├── Views      — twin-wide dashboards
  └── Scenarios
        └── Scenario (e.g., "Baseline Growth")
              ├── Views  — dashboards for this scenario
              ├── Agents — the live elements (one per real-world thing)
              └── Links  — connections between agents

Each scenario is independent — what you change in one doesn't affect the others. Give the twin a calendar and its timesteps become real dates: weeks, months, quarters (see Putting a Twin on a Calendar).

Try it live

The intro model includes a working example: the Limits to Growth archetype. It shows how a growing population eventually saturates against a finite carrying capacity. Drag the slider and watch the system reach its limit sooner or later — that's a live twin in action.

Next steps

Creating Twins and Scenarios — set up your first twin
Adding Agents to a Scenario — bring nodes to life on a simulation diagram
Entering Time Series Data — use the spreadsheet to record values
Playing Through a Simulation Step by Step — turn a twin into a game
Putting a Twin on a Calendar — work with real dates