Artificial Intelligence

AI in Rapise

We have integrated Inflectra.ai capabilities directly into Rapise, allowing testers to harness the power of AI directly inside Rapise. The AI engine embedded in Rapise is fully trained on Rapise's Global Object API, and understands the testing frameworks your team has created in Rapise, including your modules and object repositories. This knowledge enables the generation of precise and specialized responses rather than just generic answers. It also avoids the pitfalls of hallucinations found in generic AI tools.

Rapise’s AI features include RVL and JavaScript code generation, as well as data generation functionalities. When connected to Inflectra.ai, Rapise, has access to a large range of models from Amazon and Anthropic, including multimodal ones that can accept image input. The AI workflow in Rapise also supports incremental improvements in code generation quality by allowing users to save positive and negative examples for use in subsequent prompts.

Key Benefits of AI in Rapise

Accelerated Script Development: Imagine writing human-readable test scenarios and having AI translate them into fully functional automation scripts in no time. That’s right—what used to take hours can now be done in minutes, speeding up your development cycles and minimizing manual effort.

Simplified Script Maintenance: As your application grows and evolves, keeping test scripts up-to-date is a challenge. Rapise’s AI features help keep those scripts in sync with your app, offering guidance to ensure your test suite stands the test of time.

Synthetic Test Data Generation: Need a variety of realistic data to put your app through its paces? Rapise has you covered with AI-driven test data generation, which creates diverse datasets on demand, improving both coverage and accuracy.

Intelligent Assistance: Whether you're an automation expert or just getting started, Rapise's AI gives you a boost, offering contextual advice, tips, and support through every phase of testing.

Technical Highlights

Flexible AI Integration: You’re in control. Rapise’s AI features can be customized and extended to fit your exact needs, ensuring that your automation strategy stays as agile as your development process.

Seamless Collaboration: Our AI bridges the gap between human testers and automation tools, making it easier for teams to work together, whether you’re writing code or running manual tests.

AI-Powered Features in Rapise

What makes this really exciting is that all of these AI capabilities are built directly into Rapise—no need to jump between tools or platforms. Whether you're generating RVL or JavaScript code, or creating test data, everything is managed within the Rapise interface.

Our AI features are designed to work hand-in-hand with Rapise’s Global Object API, along with any open framework, its modules, and object repositories in use. This gives the AI the context it needs to deliver accurate, specialized responses, which means you’re getting tailored solutions, not just generic advice. By default, Rapise uses Inflectra.ai as its AI engine, but it can also use a customer’s own instances of Amazon Bedrock, Azure, and OpenAI models if preferred. The AI in Rapise learns over time, saving positive and negative examples to enhance the quality of future code and test generation.

AI Commands: Turning Test Steps into Code

One of the key features in Rapise is the AI Command. Simply describe your test step in plain text, and Rapise's AI converts that description into executable JavaScript. And the best part? Once the code is generated, it can run independently of AI. Modify the test step description, and Rapise automatically regenerates the corresponding code.

It's a seamless and powerful way to generate test scripts. When using AI Commands, we recommend a streamlined workflow: start by defining Page Objects with high-level actions (like Login, AddRecord, DeleteRecord). Next, add any test case-specific objects to your Object Repository. Finally, use AI Commands to turn those actions into code. With this approach, the AI generates code that you can easily review, edit, and execute—giving you a complete test case without all the hassle.

Two Interfaces, One Powerful AI

Rapise makes it easy to interact with AI through two main interfaces: the AI Dashboard and the AI Panel. These give you full control over how AI is integrated into your test creation process, ensuring you always have the insights and tools you need at your fingertips.

AI Dashboard

The AI Dashboard allows you to configure AI options, chat with the AI, and view your chat history.

AI Panel

The AI Panel is linked to the active RVL document and is used to translate AI Commands into executable JavaScript code.

With Rapise’s AI capabilities, you’re not just automating tests—you’re revolutionizing how your team approaches testing altogether. Whether you’re creating, maintaining, or running tests, Rapise gives you the power of AI to streamline and elevate every aspect of your testing journey.

Write Tests in Natural Language with AI Translator

AI Translator lets testers describe automation steps in natural language and have Rapise translate those instructions into executable SmartActions.

Instead of having to express every interaction in automation-specific syntax, a tester can write instructions such as “Log in as the standard user,” “Search for the customer record,” or “Verify that the order appears in the results.” AI Translator interprets the intent of those instructions and generates the automation needed to perform them.

Importantly, AI Translator is not simply generating generic automation code. It works with Rapise's existing testing framework, including Page Objects, Modules, learned objects, and the Object Repository. It translates natural-language instructions into SmartActions, giving the resulting automation access to the same self-healing capabilities that help Rapise tests adapt when applications change.

Advanced Visual Testing Capabilities with AI Tester

Rapise makes visual testing easy and straightforward with its AiTester functionality. AiTester enhances the testing process by introducing advanced visual testing capabilities, simulating the keen observational skills of a manual tester. This feature enables automated detection of visual discrepancies, such as layout shifts, missing elements, or subtle changes in design, that traditional functional tests might overlook.

With AiTester, you can send text and image-augmented queries to AI models hosted by Amazon Bedrock, OpenAI, and Azure OpenAI. Example applications include generating data, performing image-based verifications (e.g., detecting discrepancies or counting visual elements), and more.

By incorporating AI-powered analysis, AiTester ensures a more thorough review of the user interface, identifying issues with the precision and adaptability of human judgment, all while maintaining the speed and consistency of automated execution.

For more details on using AiTester in your environment, and examples of the different use cases, please review KB883 - AiTester Public Module, but here are some of the most common uses:

Comparing Images

This functionality will automatically analyze the image from the original test recording with that at playback and describe any differences found. In the example below, you can see that the testing engine has found the various differences between the two pictures (number in the clouds, sun vs. butterfly in the sky, color of the umbrella, etc.)

Assess the State of an Application

The general image query lets you ask open-ended questions about the image in question. In this example, we are asking the engine if the tester has already logged in or not (spoiler alert, the answer is No):

Comparing Screenshots

The same image comparison feature we illustrated above can also be used with the specific task of comparing screenshots. In this case we are able to get discrete, concrete differences between the two screenshots. They are in fact two versions of the same application, one written in React, the other in Angular.

Performing AI-based Assertions

Finally, you can use the AI Tester to ask opinionated questions such as whether a page looks "OK" and have the AI give its reasoning as to why (or why not!). In this example, all of the UI elements are clearly visible and district from each other.

Whereas by comparison, the following screen has at least one overlapping element:

Analyze Test Results with AI TestRunner

Rapise includes a powerful test orchestration engine called TestRunner. In addition to being able to schedule and orchestrate multiple parallel runs of tests at scale, it also leverages Inflectra.ai to perform AI-powered root cause analysis, and also to aggregate test execution data for longitudinal reporting.

Analyzing Test Failures

Using Inflectra.ai and the TestRunner module, you can analyze large volumes of test results and perform complex AI queries to look for trends and/or patterns:

In this example, we asked the AI to discover which test cases had at least one non-pass status and draw all statuses into a plot over a time series. In this plot, each test case is a horizontal sequence of dots.

Identifying Flaky Tests

Another common task that AI TestRunner can help us with is looking for the top flaky tests in the test data and visualize them to look for patterns.

In this example, we queried the last 10,000 automated test runs in Spira for a specific application. We specified that we wanted to find the top 20 most flaky tests. The results are visualized in an plot graph that can be uploaded to Spira for enterprise reporting and historical archiving.

Agentic Workflows with MCP Chat

MCP Chat introduces an agentic conversational interface that connects AI models with the tools and information available through Model Context Protocol (MCP) servers. In Rapise the chat can work with the Rapise and Spira MCP Servers as well as other third-party MCP servers such as GitHub and GitLab, creating a bridge between test automation, test management and CI/CD pipelines.

Instead of navigating through multiple screens or manually invoking individual commands, users can describe what they want to accomplish in normal language.

For example, a tester might ask Rapise to find the tests associated with a particular area of the application, inspect an existing test case, create a new test, execute automation, or investigate the outcome of a test run.

The AI agent determines which available MCP tools it needs to use and can orchestrate multiple tool calls to accomplish the requested task. Rapise can also present the proposed plan before execution, allowing users to review what the agent intends to do before approving actions.

This is an important distinction between conventional AI chat and agentic AI. The assistant is not limited to answering a question based on information contained in its language model. Through MCP, it can interact with the testing environment itself.

Autonomous Exploratory Testing with AI Robot

In addition to the AI features above, Rapise also includes the powerful Agentic AI AiRobot feature. This feature allows Rapise to perform automated exploratory testing using Agentic AI workflows. This also allows execution of manual tests directly with minimal or no conversion from the manual test case.