Skip to main content
Population
We are currently offering to build custom population models for API users. Send a request here to get started.
Populations are the core of primitive of the Semilattice API.

​
What is a population?

A population is a model that represents a specific group of people in reality. When you call the API with a population_id, you’re telling Semilattice which group to simulate.

​
Target population

Population models have a reality_target metadata field which describes the group of people in reality the model is intended to represent. When choosing a population model, select the one that best matches your target audience or user profile.

A Population modelling global, professional devops specialists

Populations range from broad user profiles to highly specific niches.

​
Seed data

Populations are built using tabular question-answer data (such as survey data) from the target audience. This seed data determines:
  • Prediction accuracy: How well the model predicts the target audience
  • Subject expertise: What topics the population can predict accurately

Seed data define population models

Learn more about how populations are created in the seed data section.

​
Accuracy evaluation

Every population model, once tested, has evaluation metrics that help you understand prediction accuracy:

Accuracy metrics help assess prediction confidence

These metrics appear in API responses and help you:
  • Choose the right population for your use case
  • Set appropriate confidence levels for your predictions
  • Understand the reliability of specific results
Learn more about these metrics and how to interpret them in the Accuracy Evaluation guide.

​
Simulation engine

Each population uses a specific version of Semilattice’s simulation engine. This ensures:
  • Consistency: All predictions use the same model version
  • Accuracy: Evaluation metrics reflect real prediction performance
  • Reliability: Results are reproducible

Populations use consistent simulation engines

The simulation_engine field in API responses tells you which engine was used. All public population models use the answers-1 simulation engine.

​
Public vs. custom populations

When making API calls, you can access:
  • Public populations: Available to all users with valid API keys
  • Custom populations: Only accessible to the owner’s API key
  • Can be used by any authenticated API user
  • Ideal for common user profiles (“US Adults”, “Developers”)
  • IDs can be shared in documentation and examples
  • Only accessible with the owner’s API key
  • Perfect for proprietary customer segments
  • IDs should be kept secure like API keys, though only owners of a custom population can use it for simulations

​
View available population models

To use a population in your API calls, you need its ID. You can view which population models are available for simulation in two ways:
  1. Call the REST API GET /populations or the SDK populations.list methods:
response = semilattice.populations.list()
populations = response.data
  1. Visit the populations page in the dashboard:

Click to copy a population's ID from the dashboard