Skip to main content
Communities in Bluejay are collections of Digital Humans designed to help you organize, manage, and reuse realistic test scenarios across multiple agents and experiments. They act as an efficient way to group related Digital Humans — such as those focused on specific customer types, languages, or goals — so you can quickly run consistent sets of tests whenever and wherever you need.

​
What Is a Community?

A Community is a named and described group of Digital Humans that share a common purpose or testing focus. Communities allow you to manage your testing personas in a structured, repeatable way—ideal for running the same set of Digital Humans across different agents or evaluation cycles. For example, you might create a “Spanish Support Callers” community to test how different agents handle multilingual customer interactions, or a “Difficult Customers” community to stress-test escalation and empathy handling.

​
How Bluejay Constructs Communities

Communities are built from existing Digital Humans within your workspace. Each Community includes:
  • Name
    • A clear, descriptive title to identify the testing focus.
  • Description
    • Context about the scenarios or traits that unify the group.
  • Digital Humans
    • A curated set of virtual customers sharing thematic or behavioral similarities.
This structure ensures you can maintain organized testing setups while allowing your teams to iterate on consistent, comparable benchmarks.

​
Importance of Communities

Communities make testing and observability more efficient by enabling:
  • Repeatability
    • Run the same group of Digital Humans across multiple agents for consistent evaluation.
  • Organization
    • Keep your Digital Humans categorized by purpose, goal, or customer type.
  • Scalability
    • Reuse entire communities in new simulations, avoiding repetitive setup work.
  • Comparability
    • Benchmark agent performance across standardized human groups.

​
Best Practices

  • Clear Naming
    • Use concise, descriptive titles (e.g., “Healthcare Inbound Patients” or “Upset Callers”).
  • Detailed Descriptions
    • Write meaningful descriptions explaining the testing goals or traits that define the group.
  • Regular Updates
    • Refresh your Communities as your Digital Humans evolve to reflect new business needs or testing priorities.
  • Cross-Agent Use
    • Leverage the same Community across agents to identify relative strengths and weaknesses in performance.

​
Example Community

Name: "Spanish Billing Support"

Description: "A set of frustrated Spanish-speaking customers focused on billing disputes, designed to test multilingual support and tension-handling ability."

Digital Humans:
- María González – Frustrated, Fast-paced, Latin American accent
- Jorge Ruiz – Calm but firm, Neutral tone, Spain accent
- Ana López – Impatient, Loud, Latin American accent

Scenario Goals:
- Assess multilingual handling under stress
- Measure response accuracy and empathy
- Benchmark escalation performance across agents