Mabl, as an AI-powered low-code tool, has been recognized for many years as a leader in the domain of tools leveraging various artificial intelligence solutions, from Machine Learning (ML) to generative AI (GenAI), and its current entry into the domain of AI Agents and the MCP protocol.
These topics were also a key part of this year’s Mabl conference, which I would like to share from the perspective of a speaker.
Mabl Experience 25 Conference
The Mabl Experience 25 Conference attracted twice as many attendees as last year. The event took place in a unique location, offering a picturesque view of the Charles River and Boston Harbor.
Many people may be wondering how I ended up there.
- Firstly, Mabl is a partner of Sii, which I will mention later.
- Secondly, thanks to my involvement in this partnership almost from the very beginning, I can boast the role of ambassador for this tool.
- And the most important thing is my participation as a speaker at this year’s event.
The World of AI Agents and MCP
Agents and MCP (Model Context Protocol) were topics repeatedly discussed throughout 2025, so it was no surprise to see them featured at this year’s conference. Mabl has not only implemented the MCP protocol, but also introduced and refined its own agents. These two functionalities work exceptionally well together.
Using natural language, we can leverage MCP to create test cases, for example, based on various artifacts from Jira.
Mabl’s new functionality allows not only the creation of individual tests but also the generation of multiple tests directly from MCP. Simpler test cases are created in the cloud, while more complex ones (likely requiring verification by a tester) are executed within the tool. A tester can pause an agent’s work at any time to correct any errors. Moreover, before implementing a test case, we have the opportunity to iterate multiple times during the planning phase, almost as if we were collaborating with a team member.
My session at Mabl Experience 25
During my talk, I focused on a key topic: managing both our tests and our teams. Practices like TestOps play an essential role here, helping us bring order to the chaos and optimize our workflows. It’s precisely the Agents I mentioned earlier that are now challenging this status quo.
When it comes to the main reason why many human-driven processes fail, most of us would probably agree on one thing: it’s communication.
Misunderstandings often start with the collaboration between the PM, developers, and testers. Adding AI into this communication triangle makes things even more complex. Think of AI as a fleet of messengers: not all of them deliver information correctly.
As you can see, new practices and behaviors are necessary to manage this effectively.
Challenges in a world with AI Agents
In exploring the new world with AI Agents, I identified three key challenges:
- Test Library Overgrowth. Agents can be triggered by testers, developers, and business stakeholders, which may lead to duplicate or unnecessary test cases.
- Talent Gap. If Agents can perform Junior-level tasks, what does this mean for the development of future Seniors and Leaders?
- Bottleneck Shift. It is much easier to scale Agents for developers’ tasks than to hire additional testers. However, when verification of their results is required, the bottleneck shifts from test creation to test validation.
A clear message from my session and others at the conference was that human involvement remains crucial to ensure compliance with standards and technical or legal requirements. By delegating test execution to Agents, we can focus more on preparing test cases, workflows, and validation. Roles evolve from QA Engineer to QA Auditor, Conductor, or Architect.
This also presents opportunities for future QA professionals: auditing Agents’ work can help them gain the skills needed to become tomorrow’s leaders, much like how PR reviews help sharpen skills today.
Partnership
Beyond my presentation, as Mabl partners, we were honored with an award for one of the year’s best partnerships, which I had the pleasure of receiving with my colleague Julita Machelska, Account Executive at Sii.
Summary
The conference was both successful and inspiring. Being at the heart of IT transformation every day, with AI-powered tools evolving rapidly, is truly exciting. Witnessing one of these tools develop firsthand, especially in such a prestigious location and among so many passionate experts, was a truly unique experience.
I’m already looking forward to the challenges 2026 will bring. How will these tools evolve next? Will we see another breakthrough like MCP, which appeared suddenly and is now everywhere?
I hope to share all of Mabl’s functionalities soon. Maybe even giving us a chance to meet again at the next conference!
Leave a comment