Why Use Automation Intelligence (Eggplant Test)
AI software testing built for the real world, not the demo.
Most test automation tools were built for simple web apps. Modern enterprise software isn't simple. It works across browsers, mobile, APIs, mainframes, virtualized desktops, and regulated environments, where traditional scripted tests often break when the UI or environment changes.
Eggplant Test is different. Instead of relying on large libraries of manually scripted paths, your QA team models how the application is used. The AI then generates and runs tests through that model, with scripts handling the UI and API actions.
You get broader end-to-end coverage. Less maintenance. Faster feedback. Evidence-based release decisions instead of crossed fingers.
From Requirements to Test Cases in Minutes
Keysight Generator reads your requirements file and produces Gherkin scenarios and manual test cases automatically, without manual test design effort. It runs on-premises, keeping your requirements data inside your environment.
Engineering Applications
Point it at your engineering stack.
Automation Intelligence (Eggplant Test) tests at the UI level with image recognition, so it runs on any engineering application that renders on a screen, SAP, Teamcenter, Windchill, NX, 3DEXPERIENCE, Creo, and the integrations between them. No API access, no vendor-specific connectors, no code changes. A single test can walk one workflow across all of them at once, the way engineering data moves from design through production.
Faster Time to ValueAI test automation starts once the workflow model and execution assets are in place. Less manual path scripting, less duplicated test effort.
Detect Issues EarlyAI explores paths a human tester might not think to try, helping teams catch defects earlier in the lifecycle.
Cover Critical User JourneysValidate user journeys across web, mobile, API, Citrix, desktop, embedded, and legacy systems from one automation approach.
Test the ComplexIf a user can interact with it visually, Eggplant Test can often test it without access to the application code. That includes secured systems, virtualized desktops, and legacy infrastructure your scripted tools can't touch.
See It in Action
Eggplant Test works from a model of how people move through your application, then generates and runs thousands of paths, including the combinations no one would script by hand. It reads the screen the way a person does, so it reaches interfaces DOM- and code-based tools can't: packaged software, legacy systems, embedded UIs, virtual desktops. And one test can carry the whole journey, through the UI, then into the database to confirm the data landed.
Take the tour and watch it find the paths your scripts miss.
Core Capabilities
AI-driven testing across web, desktop, mobile, API, Citrix, and complex multi-system environments. Validate complete user journeys with results, screenshots, logs, and analytics.
Test the UI the way a real user sees it with image and text recognition, not DOM selectors. Tests remain stable regardless of what changes in the front-end stack.
AI generates and prioritizes paths through the model, including combinations no human would script. Update the model when the app changes, then the AI re-explores.
Validate API behavior and end-to-end workflows in one platform. Combine API checks with UI automation to catch integration bugs before they break the user experience.
Connect to CI/CD through public APIs and integrations including Jenkins and Bamboo. Run automated tests as part of any build or release process.
Reuse the same automation approach across browsers, operating systems, devices, and remote environments, without depending on DOM locators or object models.
Use Case
Use Case
Use Case
Use Case
Resources to Get You Started
For most automation teams, maintaining test scripts costs more time than writing them. Every UI change, theme update, or redesign breaks pixel-matched references, and engineers spend hours fixing scripts instead of extending coverage. Find by Description changes that equation entirely, using AI to locate UI elements from plain-English descriptions so tests stay current without maintenance.
Blogs
More Resources
eBooks
The Ultimate AI Testing Playbook
白皮书
Testing Trends in 2025
报告
Transforming Quality Assurance and Testing with Agentic AI
eBooks
The QA Engineer’s Guide to Mobile App Testing
报告
Complexity Overload and Bottleneck Struggles: The Hidden Costs of Manual PLM Testing
eBooks
NASA’s Gold Standard Testing Methods and Software Test Automation
Have Questions? We've Got Answers
Eggplant Test is an AI-powered test automation platform from Keysight for end-to-end software testing across web, mobile, API, desktop, and complex systems like Citrix. Teams model user journeys and system states; Eggplant Test generates and runs paths through those models using functional execution assets. It is used by enterprises in regulated and complex environments to expand coverage, reduce maintenance, and improve release confidence.
The AI explores model-based test paths that a human may not think to write, including edge cases, regression risks, and unusual user behaviors. Defects can surface earlier in the cycle when they cost less to fix. Coverage expands without requiring teams to manually script every path. Customers can reduce manual testing and maintenance effort, depending on scope and implementation maturity.
Eggplant Test can test web, desktop, mobile, API, Citrix/VDI, mainframe, kiosk, embedded, and proprietary GUI environments where visual or remote access is available. It supports Windows, macOS, Linux, iOS, and Android targets through supported connections and integrations, and API tests can be included alongside UI workflows.
By reducing the need to script every path by hand. When the application changes, teams update the model and reusable snippets, then Eggplant Test re-explores the workflows instead of forcing engineers to rewrite large sets of linear scripts. That helps teams maintain broader regression coverage with less manual rework.
Yes. Eggplant Test is designed for enterprise-scale testing, including distributed execution, centralized test asset management, CI/CD integration, results analytics, and requirements traceability. It is relevant for regulated industries including aerospace, defense, digital engineering and manufacturing, and financial services where evidence, auditability, and release confidence matter.
Scripted automation requires engineers to write each test path by hand. When the application changes, affected scripts often need to be updated. Model-based testing flips this: engineers build a model of the application's states and actions, and the AI generates and runs paths through that model. When the application changes, teams update the model or reusable snippets and the AI re-explores. This produces broader coverage with less maintenance than maintaining large sets of linear scripts.