Systems | Development | Analytics | API | Testing

MLRun 1.10, the latest version of our open source AI orchestration framework, is available today to all users. Iguazio started out as a platform to operationalize enterprise machine learning projects. Though we’ve been through quite a few waves of AI in just a short time, the underlying challenges are the same: getting from experimentation to production remains a major blocker.

We’re entering a phase where AI can draft emails, resolve tickets, summarise complex information, and occasionally present fiction as fact with equal conviction. Generative AI has become incredibly powerful, but in enterprise environments, power without precision quickly becomes a risk rather than an advantage. This is exactly where the shift is happening.

In the last ten years, software testing has advanced significantly, but today’s applications require more than just using conventional forms of automated software testing or entry-level tools that employ artificial intelligence (AI). The rise of microservice architectures, API calls, and continuous deployment has led to another category of software testing products called "Generative" AI Testing tools.

This blog expands on insights Amjadh Ifthikar shared during his WSO2Con Asia 2025 presentation. Watch the full session here.

Over the past few years, model providers have invested heavily in “guardrails”: safety layers around large language models that detect risky content, block some harmful queries, and make systems harder to jailbreak.

The EU AI Act is here, and for many enterprises, it represents a massive coordination challenge. As the world’s first comprehensive AI law, it mandates strict governance on transparency, risk management, and data quality. For platform engineers and architects, the immediate question is operational: How do we comply with these new regulations without forcing every developer to rewrite their applications?

Note: This article is the third instalment in a four-part series. Check out the first two posts here:- Why we ditched frontier AI agents and built our own- Building Bitrise’s context-aware, browser-integrated AI copilot that keeps engineers in control.

If you’ve worked in QA long enough, you’ve seen how much testing has changed inside Jira. What started as a mix of spreadsheets, manual checklists, and endless review cycles has grown into fully integrated Test Management workflows. But even with automation, some challenges never went away. Writing test cases from requirements still takes time.

Testing is a very important and necessary step in the SDLC, but most teams ignore it or don’t care much about it, while some teams spend most of their time on testing instead of building features. AI is really changing the way we write code, but most people use it mainly for writing test cases, and we still end up doing it manually. So in this blog, let’s see what AI testing is, how AI helps in testing our software, what AI tools are available, and which tools help with which part of testing.