Three ways UiPath Test Cloud is paving the way for agentic, autonomous, end-to-end QA

UiPath Test Cloud isn't just built for the testing you know today—it's built for where testing is going. Yes, it helps you automate the traditional, narrow, and rule-based tasks in testing—the stuff we've all been doing for years. But more importantly, Test Cloud gets you ready for what's next: automating the broader, messier, more dynamic, and a lot less predictable challenges in testing.

We call this new paradigm agentic testing: augmenting testers with AI agents throughout the testing lifecycle so they can extend their capabilities, achieve things that weren't possible before, and ultimately pave the way for autonomous end-to-end testing. That's agentic testing, and Test Cloud brings it to life in three ways.

1. Leverage built-in AI agents to get started fast

Test Cloud provides a broad range of built-in AI agents assisting you throughout the entire lifecycle, from test design and test automation to test execution and test management.

These AI agents are powered by UiPath Autopilot™ and fall into four key categories:

• Agentic test design: Autopilot for Test Manager empowers you to evaluate requirements (user stories) for quality aspects (clarity, completeness, consistency), generate manual tests from requirements, and identify obsolete or redundant manual tests.

• Agentic test automation: Autopilot for Studio enables you to convert text like manual tests into coded and low-code automated user interface (UI) and API tests. It can generate synthetic test data and refactor existing test automation. Autopilot also fixes validation errors, creates expressions (regular expressions), and performs fuzzy, non-deterministic verifications. You can even collaborate with Autopilot conversationally to create, update, and improve your test automation (vibe-coding).

• Agentic test execution: Autopilot repairs broken automated tests at run time through its GenAI-powered self-healing capabilities. It identifies the root cause of failures and adjusts selectors or logic so tests continue running without disruption. Autopilot can also execute manual tests autonomously with no coding or scripting required. It reads manual test steps, interprets the intent, interacts with the application, captures evidence, and records the results.

• Agentic test management: Autopilot for Test Manager gives actionable insights into test results, import manual tests, lets you search test projects using natural language. You can also chat with Autopilot for Test Manager to get guided support across all phases—from requirements creation and test generation to reporting—making it your always-on digital testing companion.

2. Create custom AI agents to build exactly what you need

Test Cloud gives you everything you need to build and evaluate your own custom autonomous or conversational AI agents, giving you the flexibility to build exactly what you need, when you need it, your way. This comes to life through Agent Builder in UiPath Studio.

Agent Builder lets you define natural language prompts that define the agent's goals, roles, and responsibilities. You can use active and long-term memory through our context grounding capabilities to inform how the agent thinks and plans, all grounded in your enterprise context. You can equip agents with tools like coded or low-code UI, API, or database automations, Model Context Protocol (MCP) tools you either build in Test Cloud or bring into it, and other first- and third-party agents. You can also include people in the loop for guidance (providing extra information, critical decision making) using interfaces created in UiPath Apps and managed through Action Center. And to top it off, Agent Builder in Studio not only helps you build agents but also test them with its rich evaluation capabilities.

3. Orchestrate agentic testing to make it all work end-to-end

UiPath Maestro™ is the orchestrator that turns AI agents, automations, and people into a single coherent testing flow for a unified symphony of agentic testing. It's the layer that makes every piece work together to deliver end-to-end autonomy across the testing lifecycle, rather than leaving you with isolated pockets of intelligence from individual built-in or custom AI agents.

Imagine you're running a full-scale regression suite across multiple environments, stages, and tools like Playwright, Postman, Appium, and UiPath Test Cloud. A built-in AI agent generates test cases from new requirements like user stories. Another AI agent automates those manual tests across web, API, mobile, and desktop apps like SAP®, Oracle, and EPIC. Another one generates test data for the entire execution. During execution, an AI agent for self-healing automation steps in to repair automated tests that break. A separate AI agent reviews failed tests and proposes fixes, while another AI agent surfaces failure patterns and produces stakeholder-ready reports.

But testing needs more than AI agents. You may need traditional UI, API, or database automations to read config files to spin up, tear down, or reset test environments (e.g., development, staging), or notify stakeholders via email, Slack, or Microsoft Teams about the test results. And in some moments, you need human testers in the loop, for example, for reviewing and approving how AI agents repaired automated tests before they become the new baseline for the next execution.

Maestro brings it all together. It coordinates every AI agent, every automation, and every human into one continuous, governed flow. Without Maestro, all these intelligent components operate in silos, and you end up only automating fragments of the testing lifecycle. With Maestro, the full chain of testing activities becomes autonomous from start to end.

Overall, without Maestro, autonomy is scattered. With Maestro, it becomes a system—and only then autonomous testing becomes real.

What's next?

This is only a small glimpse of the capabilities available in Test Cloud. Watch the keynote from our FUSION 2025 conference (keynote is available in Volume 1) to see how Test Cloud helps testing teams ship faster with greater confidence.