We’re excited to officially kick off our next Ask Katalon Anything (AKA) Edition 9, our regular flagship community event, designed to bring you closer to the people building Katalon.
Event Theme
Rethinking Test Recording & Designing Agentic Workflows
This AKA is all about exploring how AI-powered test recording and agentic workflows can work for real-world testing.
You’ll get a chance to ask questions directly to:
Xuan is part of our Katalon Product team and looks after Katalon Studio (StudioAssist and Reporting). Her mission is simple: to make Studio even better, more powerful, and more efficient with every update. She also loves engaging with the Katalon community, so if you have any questions about Studio, don’t hesitate to reach out to her. You’re in great hands!
Meet Thuy Ngo, Product Manager at Katalon with a strong focus on end-to-end AI-driven workflows. Her expertise spans Web and Mobile test creation & execution. Passionate about helping teams work smarter with data and automation, Thuy is always looking for ways to bring greater insights and efficiency to the testing process.
Duy builds impactful features for Katalon Studio. Passionate about creating solutions that empower testers, he enjoys seeing how the community applies Studio to solve real challenges.
Event Details
Start date: January 27
End date: February 12
Location: Katalon Community Forum
Format: Open Q&A (dedicated threads per product)
Post your questions anytime during the event, no live session pressure, just thoughtful conversations.
1. Where to Ask
Jump into the dedicated threads below to post your questions:
can the Recording Agent help bridge the gap between exploratory/manual testing and maintainable automation, especially for teams that don’t have strong coding skills?
many AI demos look great but don’t survive real projects, how do you envision agentic workflows fitting into an existing QA process (e.g., regression suites, CI/CD and team collaboration) without creating more maintenance overhead?
Why should organisations choose Katalon Studio when we we have tools like Playwright with MCP + AI agents and that too free and zero license cost, What is the roadmap for Katalon Studio ?
When the recording agent selects locators (XPath/CSS/attributes/AI-based), can we see an explicit ranking + rationale (and a fallback chain) so we can predict flakiness before it hits CI?
Do you plan to expose a diff/audit trail for AI-generated edits (before/after steps, objects changed, locators changed) so reviewers can approve changes the same way they’d review a PR?
What is the explicit security model for agentic workflows: what actions can an agent take automatically (create/modify files, edit test objects, run executions), what requires confirmation, and how can admins enforce policy?
Do you have any plans to make katalon studio cloud where we can open it on a browser ? Hence eliminating the need to download katalon studio on our machinines?
can the Recording Agent help bridge the gap between exploratory/manual testing and maintainable automation, especially for teams that don’t have strong coding skills?
The short answer is yes—the Recording Agent is specifically designed to help manual testers transition into automation more smoothly.
Here is how it helps bridge that gap:
Low-Code Entry**:** It captures your manual interactions and converts them into test scripts automatically, which is perfect for teams without deep coding expertise.
Streamlined Workflow: You can find the full setup and workflow details in our documentation: Katalon Web Recorder Plus.
A Note on Complexity: While it simplifies the process, please keep in mind that handling highly complex website behaviors may still require some foundational automation knowledge to ensure your tests are robust and maintainable.
Many AI demos look great but don’t survive real projects, how do you envision agentic workflows fitting into an existing QA process (e.g., regression suites, CI/CD and team collaboration) without creating more maintenance overhead?
Thank you for raising such a practical concern! We believe the core principle for any successful adoption is “the right tool for the right job”. To avoid overhead, we must distinguish between where AI adds value and where it adds “noise”.
Using AI just for the sake of AI can actually increase technical debt. Here are two examples of violating the “right tool for the right job” principle (producing overhead):
Over-automating Simple Verifications: Using complex AI agents to verify static elements (like a logo or a simple “Submit” button) that are better handled by standard, lightning-fast CSS/XPath selectors.
AI for Unstable Environments: Trying to use AI to “fix” tests in an environment that has inconsistent data or frequent network outages; AI cannot fix underlying infrastructure instability and will only generate “flaky” suggestions.
Regarding your question about a cloud-based version of Katalon Studio, we currently do not have plans to move the IDE entirely to a web browser. In our experience, the highly intensive interactions required for a robust Integrated Development Environment (IDE) are often better supported by a dedicated desktop application. This choice allows us to maintain the performance and stability needed for complex automation tasks.
However, we are always looking for ways to improve our tools based on user needs. Could you share more about the specific benefits you’re looking for in a web-based IDE? For example, are you primarily looking to simplify the installation process or improve collaboration across different machines?
Understanding your specific use case helps us better evaluate our product direction. Thank you for your feedback!
can we see an explicit ranking + rationale (and a fallback chain) so we can predict flakiness before it hits CI?
Thank you for sharing such a great suggestion!
We really appreciate this feedback. Providing visibility into the ranking and fallback logic would definitely help teams better understand selector reliability before tests ever reach CI. I will make sure to share this with our Product team for further consideration.
To help us better understand your needs, could you clarify a few points?
Fallback Chain: Since the AI-driven engine selects the primary locator based on the highest confidence score, how would having visibility into the subsequent fallback chain specifically assist your workflow?
Ranking: What specific metrics or types of ranking data would be most useful for you to see when predicting potential flakiness?
It’s a topic many teams are exploring as the landscape of AI-driven automation evolves.
While tools like Playwright are powerful—especially when combined with custom AI agents—the right choice usually depends on your team’s technical stack, maturity level, and time-to-market requirements. Here are two specific scenarios where Katalon provides a distinct advantage:
Accelerated Time-to-Market for Diverse Teams**:** Playwright requires a high level of coding expertise to build and maintain a robust framework from scratch. Katalon’s “low-code” interface allows manual testers and business analysts to contribute to automation immediately, while still providing a full scripting mode for advanced engineers. This hybrid approach significantly reduces the time needed to get your first test suites running.
Unified, All-in-One Ecosystem**:** To match Katalon’s capabilities with Playwright, you often have to integrate multiple third-party tools for reporting, mobile testing, and AI-driven self-healing. Katalon provides these out of the box. For example, our AI-powered self-healing and TrueTest™ (which models real user journeys) are native features, saving your team from the high “hidden cost” of maintaining a complex, custom-built toolchain.
I hope this helps clarify the direction!
What is the roadmap for Katalon Studio ?
Currently, we do not have a public-facing roadmap document that lists specific upcoming release dates for all Katalon products. However, we are deeply committed to transparency and continuous improvement based on your feedback.
Here is how you can stay updated and even help shape that roadmap:
Feature Requests**:** We actively monitor our Feature Request category on the forum. If there is a specific capability you’d like to see, please post it there! Our product teams review these regularly to prioritize what we build next.
Product Roundups**:** Keep an eye on the Announcements section for our monthly and quarterly product roundups. These posts provide deep dives into the latest features we’ve just released, such as our recent AI-powered StudioAssist enhancements and Mobile Self-Healing capabilities.
I disagree to an extent. Katalon studio also requires coding expertise. We may say its low code but most of us turn to script mode. You can do a survey on this.
Again , Playwright+mcp+ai agents is free of cost and if i have to argue on this , we really dont need coding experience and that too free of cost
So , again , why should organisations choose KATALON IDE (Not talking about other prodcucts like trutest or testops).
We may say its low code but most of us turn to script mode
Thanks for bringing this up. I apologize if my previous message wasn’t clear, but I did want to highlight that while Katalon is low-code, specific scenarios do still require coding expertise.
Again , Playwright+mcp+ai agents is free of cost and if i have to argue on this , we really dont need coding experience and that too free of cost
“Cost” is a broad term, and its impact varies depending on who you ask—whether it’s a team member, a manager, or a product owner.
When comparing tools, it’s important to look beyond the price tag:
Open-source vs. Commercial: While open-source solutions may have no initial “usage cost,” they often carry a “cost of delay.” This happens when a team has to spend significant time building and maintaining an end-to-end framework from scratch.
Maintenance vs. Innovation**:** Think of it like version control. As a developer, I’d rather pay for a service like GitHub or GitLab than spend my time manually setting up and maintaining a private Git server. Paying for a platform allows me to focus on writing code rather than managing infrastructure.
The Right Tool for the Task**:** Ultimately, it is up to each team to choose the tool that best fits their specific needs and constraints.
For individual users or those just getting started, Katalon also offers a Free edition of Studio, which is perfect for a simple, “up-and-running” automation experience.
I think the conversion could be more useful when we have a more concrete context. Could you share a bit more about your specific situation or any constraints you are facing in your testing process/tool vendor ? Understanding your current setup will help us better address the specific challenges you’re encountering beyond just the financial aspect.
@xuan.tran How do agentic workflows work in a real-world QA team where testers handle mixed tasks like manual testing, automation, UI validation, and bug analysis?
Can agentic workflows assist in deciding what to automate vs what to keep manual?
@Slowey So as @duy.lam already shared, by default, each tool call will ask for user permission before running, therefore you can control all actions like read, create, and update files. Delete action is not supported yet. And all of these is managed at the user level. From Admin side, they can only control whether AI is enabled in this Account or not.
Event Theme
Event Details
1. Where to Ask
2. Quick Guidelines
3. Rewards & Lucky DrawReady to start? Drop your questions in the links above. If you have general questions about how the event works, reply here, we’re happy to help!
