Core Concepts

Understanding the fundamental building blocks of Synthetic Users.

Overview

Synthetic Users helps you conduct user research at scale using AI-powered synthetic personas. This guide explains the key concepts and how they work together.

Workspaces

A Workspace is your top-level organizational unit. It contains all your projects, team members, and billing information. You can invite team members to collaborate within a workspace.

Projects

Projects organize your research initiatives within a workspace. Each project can contain multiple studies, audiences, and synthetic users. Projects help you:

• Organize research by product, feature, or initiative
• Manage access and permissions for specific research efforts
• Track progress and insights for related studies

Create projects programmatically using the Projects API.

Audiences

An Audience defines the target demographic or user segment for your research. Audiences specify:

• Demographics: Age, gender, location, income, education
• Psychographics: Interests, values, lifestyle
• Behaviors: Usage patterns, preferences, pain points
• Context: Job roles, goals, challenges

You can create audiences in several ways:

• Manual Creation: Use create_audience to define specific attributes
• AI Generation: Use generate_audience to auto-generate diverse personas from a description
• Extension: Use extend_audience to expand existing audiences with more demographic variety

Synthetic Users

Synthetic Users are AI-powered personas that simulate real users from your target audience. Each synthetic user has:

• Identity: Name, age, background, demographics
• Personality: Traits, communication style, perspectives
• Context: Goals, challenges, experiences relevant to your research
• Consistency: Maintains character across multiple interactions

Synthetic users are generated from audience definitions and can participate in interviews, surveys, and research studies. They provide realistic responses based on their persona attributes.

View generated users in an audience via the Audiences API.

Studies

A Study is a research investigation that tests hypotheses, validates concepts, or gathers insights. Studies connect:

• Problems: Research questions you want to answer
• Solutions: Concepts, features, or designs to validate
• Audiences: Target user segments to interview
• Interviews: Conversations with synthetic users

Create studies using create_study, then run interviews with interview.

Study Components

Problems define the research questions or user challenges you're investigating. They provide context for your study.

Solutions represent the concepts, features, or ideas you're validating with users. Studies can test multiple solutions to compare approaches.

Concepts are broader themes or ideas that span across problems and solutions. They help organize insights across multiple studies.

Interviews

Interviews are conversational interactions between researchers and synthetic users. The AI conducts structured or semi-structured interviews based on your study goals.

Key features:

• Dynamic Follow-ups: Use interview_follow_up to ask additional questions
• Contextual Responses: Synthetic users respond based on their persona and the study context
• Transcripts: Full conversation history for analysis
• Insights Extraction: Automatic identification of key themes and patterns

Use study_follow_up to ask questions across all interviews in a study.

Summaries and Analysis

After conducting interviews, Synthetic Users generates comprehensive analysis:

• Summaries: Use generate_summary to create AI-powered insights from interview data
• Knowledge Graphs: Use generate_knowledge_graph to visualize relationships between concepts, problems, and solutions
• Reports: Generate professional documents with create_report
• Follow-up Analysis: Use summary_follow_up to dig deeper into specific insights

Workflow Example

A typical research workflow:

1. Set Up: Create workspace → Create project
2. Define Audience: Create audience → Generate synthetic users
3. Design Study: Create study → Link problems and solutions
4. Conduct Research: Run interviews → Ask follow-up questions
5. Analyze Results: Generate summary → Create knowledge graph → Generate report
6. Iterate: Ask follow-up questions → Regenerate interviews → Extend analysis

Files and Context

You can attach files (documents, images, research briefs) to provide context for:

• Audience generation (to inform persona creation)
• Studies (to ground interviews in specific materials)
• Problems and solutions (to provide detailed specifications)

Use create_file and upload_file_proxy to add context documents.

Real-time Updates

Track research progress in real-time:

• Event Streaming: Use stream_events to receive server-sent events as interviews progress
• Status Monitoring: Check study and interview status to coordinate workflows

Next Steps

• Conduct Your First Study - Step-by-step walkthrough
• Generate Reports and Insights - Analysis and export
• API Reference - Complete endpoint documentation