ML & AI in DevOps: Experts Predict 7 Big Changes

Smart technology is changing many facets of our everyday life — and the same goes for DevOps. Machine learning (ML) and AI in DevOps will alter how we code, how we test, and the productivity of our DevOps processes.

To learn what big changes we can expect in the coming years, I sat down with some of the experts featured in my latest book, Accelerating Software Quality: Machine Learning and Artificial Intelligence in the Age of DevOps. We discussed the transformations we expect to see over the next few years thanks to advancements in ML DevOps and AI DevOps technologies.

Keep reading to hear about the biggest changes to expect in DevOps.

7 Ways AI DevOps & ML DevOps Will Change The Game

1. Faster Releases With Test Impact Analysis

AI and machine learning are going to have a major part in accelerating releases, especially through test impact analysis, where smart technologies will actually have the ability to actually understand which test needs to be executed. Today, there’s a lot of regression testing that’s done on a very high frequency. AI and machine learning will expedite the release cycles of these tests, and eliminate most of the tests for code changes.

More teams are adopting test impact analysis technology. Right now it's in the initial phases of the adoption. AI and machine learning will improve as adoption increases.

2. Better Insights Into Quality Risks

The second area AI DevOps and ML DevOps will see benefits is in quality risks, or quality governance. This is the ability to really understand the quality risk of an action, like before you release or merge code.

Teams will be able to answer questions such as these ahead of time: Is it dangerous? Can you release it or not related based on various parameters? Was it tested? Where was it tested? Is it used in production? Who developed it? What's the history of it?

Teams will be able to combine information from Gate Ops, APM, code changes, defects, and from the testing of course, and really understand the quality risks and action that should be done based on those insights. 

3. Preference of Commercial Tools Over Open Source

A few years ago, everyone was passionate about open source. But in the last few years, there’s been a shift in mindset toward commercial tools. The reasoning behind this is productivity and value. Companies are thinking about the value and the productivity and less about whether tools are open source or commercial.

4. Big Investments in Bots

We will see a lot of evolvement in bots and virtual assistants based on artificial intelligence and machine learning in the upcoming years, starting with customer interactions, which will mainly be handled by bots. Right now, many enterprises are already spending more money on the development of chat bots. And this is pretty understandable if you consider the potential cost savings from using bot conversations. Companies will soon be saving billions on chatbots in the US.

So in short, what we see is that bots based on AI and machine learning will take a place in our life and they need some serious testing and training.

5. More Research On AI Racial Bias

In the near future, we’ll see an increase in AI safety research. Current AI-based systems do not pay attention to the diversity and the cultural aspects of human interaction with these systems.

Looking ahead, there will be more research and emphasis on this, and there will be systems in place where the current and upcoming AI based systems will not be racially biased. It’s going to be safer for humans to interact with and use these systems and production rates.

6. Growth of Low Code Dev Platforms

Low code development platforms are going to be really, really huge. We currently see customer demands exponentially increasing and developers have to churn out features very quickly to meet these growing customer demands. We already have existing AI based IDEs, like Intelicode, but moving forward they're going to be more AI-based IDEs and frameworks to enable developers to meet these growing customer demands. And they could push cutting-edge features as soon as possible.

7. Transforming DevOps With AIOps & TestOps

AIOps is more or less the next level of DevOps services as we know them right now. With all the routes in operations, such as infrastructure management, DevOps is getting more and more complex, with an increased amount of data. DevOps will require the power of AI and machine learning in order to keep up with the dynamic, constantly changing landscape.

Additionally, we’ll see the emergence of TestOps. Up to this point, QA has really been focused on automating tests, but what about automating test operations? If you think about what's involved in test operations, it starts from the artifact being available in the CI/CD pipeline all the way to automating tests and writing bug reports. That's the entire test operations process. What we will see is the emergence of an adjacent field called robotic process automation and intelligent automation. And the idea here is that you use bots to automate a lot of manual processes.

Bottom Line

Smart technologies powered by AI and ML will transform DevOps as we know it — and these seven trends are just the beginning of more changes to come.

Hear more about how AI DevOps and ML Devops will alter the testing landscape. Watch the rest of this expert panel discussion on demand, where DevOps and testing experts come together to discuss the future of smart technology and DevOps.

Watch "What’s Next in DevOps With AI/ML?" Panel

Related Content

• 5 Types of ML & AI Testing Tools
• 6 New Types of Software Defects in the Era of AI & ML
• AI Testing 101: Practical Tips for Testing AI Systems