Skip to main content
ExploreExplore
All use cases

SOLUTIONS / Population simulation

Population simulation

Synthetic personas with different backgrounds and preferences explore hypotheses, and real-world feedback keeps calibrating the simulation.

Explore
Built forUser research teamsProduct strategy teamsSocial science researchers

01Concept exploration

Explore perspectives across a population.

Compare how synthetic personas with different backgrounds respond to the same proposal. Use group patterns and individual reasons to inform research with real people.

Input
Product concept, target personas, research questions
Output
Distribution of group preferences and individual reasons
View workflow
Different preferences, the same questionSample data · Not a prediction

“I care more about the long-term cost of using it.”

The group distribution helps surface questions; individual reasons help explain the differences.

02Scenario testing

One change. Different responses.

Keep the persona sample and the questions the same, and compare answers under two conditions. Track who changed their mind, and the reasons they gave.

Input
Baseline proposal, intervention conditions, fixed persona sample
Output
Condition comparison and changes in judgment
View workflow
Same group, two conditionsSample data · Not a prediction

Start with the baseline responses.

With everything else held constant, look back at the different reasons the same persona gave.

03Group interaction

See how opinions spread.

Follow how opinions change as personas exchange information. Review their initial views, what they learned, and their subsequent choices.

Input
Persona network, initial views, interaction rules
Output
Interaction traces and records of opinion change
View workflow
How opinions change through discussionSample data · Not a prediction

“I care more about the long-term cost of using it.”

Choose a group to look back at opinions and reasons before and after information was exchanged.

IN PRACTICE

Compare reactions to two concepts

The same set of characters reads two versions of a product introduction, explains why they would or wouldn’t try it, and then takes part in a group discussion. Researchers compare choices before and after to pinpoint which information changed their judgment.

What to prepare

  • Personal profiles, interview summaries or research personas with clear sources
  • Two concepts, experimental conditions and a shared set of questions
  • Sampling method, interaction rules and dimensions to observe

Deliverables and feedback

  • Each character’s choice, reasons and later changes
  • Distributions and differences aggregated by research dimension
  • Reviewable dialogue traces and experiment configurations

CONNECTED INTELLIGENCE

Models working together.

Connect your business context, model outputs, and application feedback.

Understand people and context

OmniSoul

Forms judgments from the supplied personal background, dialogue history and current task. The simulation system manages character states and interaction records.

Input
Character profiles, conversation history, and context
Output
Context-aware responses and proposed actions

  1. Define characters and the experiment

    Record persona sources, sampling method and variable settings to create a reusable research configuration.

  2. Run controlled scenarios

    Give the same characters different concepts, keeping both individual answers and the group interaction process.

  3. Check against real feedback

    Calibrate results against actual interviews or surveys, and report distributions, disagreements and uncertainty.

FROM PILOT TO PRACTICE

Test what works.

Test representative tasks at a small scale. Record inputs, model versions, and results, then refine the solution based on what you learn.

Access and availability depend on the model and release.

Grounding in personas

Check whether each character’s judgment can be traced back to source data and experimental inputs.

Response to interventions

When only a key variable changes, compare whether choices and reasons shift in explainable ways.

External calibration

Evaluate simulation results against independent real-world feedback, recording errors and scope of validity.

Start a pilot

Tell us about your users, tasks, and success criteria. We’ll help you explore the right models.

Research and industry resources

OMNISOUL · RESEARCH PREVIEW

Request access

Tell us about your team and your research.

Research area