Tech, ExplainedCloud platforms and developer tools
Product

Test automation tools

From scripts to plain English to AI agents.

BEFORE ITAFTER ITTest automation toolsSmartBearTestsigmaAlgoShackPostQodePARTAI testing agents+COVERSScriptsPlain EnglishModel-basedSelf-healingCI/CD pipeline↗ DEVOPSstarts eachtest runAI coding agent↗ AI CODINGaddsgeneratedtestssends teststo devicesCross-browser and realdevice testingCONCEPTS IT USESAutomated testingAI coding agentsBEFORE ITAFTER ITTest automation toolsSmartBearTestsigmaAlgoShackPostQodePARTAI testing agents+COVERSScriptsPlain EnglishModel-basedSelf-healingCI/CD pipeline↗ DEVOPSstarts eachtest runAI coding agent↗ AI CODINGaddsgeneratedtestssends teststo devicesCross-browserand realdevice testingCONCEPTS IT USESAutomated testingAI coding agentsBEFORE ITCI/CD pipeline↗ DEVOPS↓ starts each test runAI coding agent↗ AI CODING↓ adds generated testsTest automation toolsSmartBearTestsigmaAlgoShackPostQodeAI testing agents+ScriptsPlain EnglishModel-basedSelf-healingAFTER IT↓ sends tests to devicesCross-browser and real device testingCONCEPTS IT USESAutomated testingAI coding agents
Tests written as scripts, in plain English, from a model of the app or by AI agents are started by the pipeline and sent to real devices on the cloud test grid.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

Test automation tools are the software in which automated tests are written, stored and run. An automated test is a small program that uses an app the way a person would, by typing, tapping and reading the screen, and then checks the result. The way tests are written has changed four times: from scripts written by programmers, to plain English, to tests generated from a model of the app, and now to AI agents.

The first generation was scripts. A test engineer writes code that finds each button and field on the screen and acts on it. This is precise, but slow to write, and it needs people who can program. It also breaks easily. If a developer renames a button, every script that looked for the old name fails, even though the app still works.

The second generation lets a tester write the steps in plain English. Consider a travel website in Gurugram that sells flight tickets. A tester who does not code writes: "Open the home page. Enter Delhi in From. Enter Goa in To. Choose 14 December. Click Search. Check that at least one flight is shown, with a price in ₹." The tool turns each line into actions on the screen. When the Search button is later renamed, the tool finds it by its position and its label, and repairs the test. This is called self-healing. Some tools can now also produce such steps from an AI model.

The third generation generates tests from a model of the application. Here, a model means a description of how the app behaves, such as its screens and the paths between them, not an AI model. The tool works out every path and writes tests that cover each one.

The fourth generation uses AI testing agents. An agent is given a goal, such as "test the refund request". It explores the app, writes the tests and runs them by itself. This is one use of AI coding agents, and a general AI coding agent now often writes tests for the code it produces as well.

However the tests are written, they run in the same places. The CI/CD pipeline starts them on every change. They run on many phones and browsers at once on a cloud test grid. Each generation has needed fewer people to write and repair the tests. That matters in India, where testing has long been done by large teams in IT services firms, and where many of these tools were built.

Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.