Builder Resources

Tutorials and templates for designing and launching surveys and interviews interactively.

Tutorials

Getting Started

Learn how to construct questions, scenarios, and agents for your surveys
View Tutorial

Running Surveys

Run surveys with AI agents and validate with humans
View Tutorial

Using Images

Create scenarios for images or other content to use with your questions
View Tutorial

Transcripts to Personas

Turn human interview transcripts into detailed personas for your surveys
View Tutorial

Piping

Add context and prior answers to your survey questions
View Tutorial

Multiple Runs

Generate multiple runs of the same AI study
View Tutorial

Sharing Content

Share, copy, and reuse content across surveys
View Tutorial

Survey Logic & Flow

Explore features for controlling survey logic and flow
View Tutorial

Builder <> EDSL

Extend your workflow in open-source EDSL code
View Tutorial

Generate & Grade a Quiz

Create a quiz, collect responses, and grade them with AI
View Tutorial

AI Interviewer

The AI interviewer conducts conversational interviews using natural language, asks follow-up questions based on respondents' answers, and captures the responses as structured, analyzable data. Instead of a fixed form, it adapts the conversation in real time, making it possible to collect richer qualitative input at scale and turn it into consistent, comparable insights for research.

AI Interviewer

Use AI to conduct conversational interviews with humans
View Tutorial

Extracting Data from Interviews

Use AI to conduct interviews, then use a model to extract details
View Tutorial

Data Labeling

AI-powered data labeling allows you to efficiently label complex datasets, evaluate qualitative content, and extract themes from unstructured text at scale by prompting models to answer structured questions about each data scenario and turning the answers into reproducible, analyzable datasets. You can also add AI agent personas when you want perspective-specific labels.

Data Labeling

Use a model to efficiently label complex datasets
View Tutorial

Data Labeling with an AI Persona

Use AI personas to label data with context and domain knowledge
View Tutorial

Data Labeling: Content Review

Use AI personas to evaluate qualitative content
View Tutorial

Quantifying Themes

Extract themes from unstructured text
View Tutorial

Validations

Validate AI results by benchmarking them against human responses, testing models with holdout data, and running robustness checks to build confidence in your insights. Use these techniques to validate your AI survey results and build confidence in your insights.

Human vs. AI Surveys

Send a survey to AI agents and humans, then compare responses
View Tutorial

Holdout Testing

Test a model's performance by asking it to predict hidden agent traits
View Tutorial

Robustness Checks

Stress test your AI survey results
View Tutorial