LLM-Assisted Test Data Generation

Overview

Autonomous enables users to generate realistic, randomized test data simply by describing what they need in plain English. You can generate anything from a simple value to a complex structured record and Autonomous handles the rest. By leveraging large language models (LLMs) to pre-generate datasets from freeform natural language prompts, this feature allows users to inject varied and relevant test data at runtime without manual scripting or setup.

Use it to quickly generate realistic, random data like email addresses, phone numbers, names, and more for testing forms, validating inputs, simulating signups, filling out shipping information, and beyond. This saves time and increases test coverage without requiring manual entry.

How it works

1. Authoring Tests

   • You write a test step using natural language that includes a data generation request.

   • Autonomous parses what you wrote, extracts the data description, and uses an LLM to generate random data
     
     
     

2. Executing Tests

   • When the test step runs, data records are randomly selected at runtime without impacting test execution speed.

.

Examples

• Generate random test data for a French fashion designer and store it in the variables first-name, last-name, email, and age

• Generate a random first name of a female that starts with "D" and ends with "K", and type it in the name field

• Randomly generate a prime number greater than 1000 and store it as prime