A survey of 2,670 software engineering professionals conducted by Tricentis, the test automation specialist, has found overwhelming support for the idea that AI technologies will transform DevOps over the next few years,
Nearly 90% of respondents said they were aware of significant potential benefits of AI-augmented DevOps, and nearly 70% rated the potential of AI-augmented testing as “extremely valuable” or “very valuable”.
The challenge in test automation is managing the massive amounts of complex data and the frequent changes to test-cases required, the survey authors say. These changes often break most automation cases, introducing extensive maintenance without the assistance of self-healing tools.
Typically, DevOps engineers use AI-augmented testing in the form of self healing test cases (cited by 49% of survey respondents who said they were using AI) and also to identify or focus on high risk testing projects (cited by 44%). Functional testing was cited by 65% as especially well-suited to AI-augmented DevOps, followed by UI and performance testing.
Improved customer experience and the reduction of the “DevOps skills gap” – making it easier for less-experienced engineers to automate testing – were seen as the two key business benefits of investment in AI-augmentation for DevOps. Whereas the key challenge firms believed held them back from AI-augmentation was the lack of data and AI skills, cited by 44%, followed by 25% lacking sufficient budget.
The survey reveals just how far many firms have yet to travel on the road to successful automation, irrelevant of the maturity of their DevOps environment. According to one DevOps leader at a US financial services firm, quoted in the survey report: “Most organisations today, including mine, still depend heavily on manual testing; mainly because traditional automated test scripts require too much time to maintain and quickly become unreliable.”
“AI-augmented DevOps is well-suited for automating specific types of testing including unit testing, functional testing, UI testing and performance testing. The big advantage is the time saved – thanks in part to the ability to create automated tests without the need for code, which saves valuable time for all teams involved.”
The Tricentis research covered 60 countries and 20 industries, where financial services accounted for 15% of respondents. 55% of respondents to the survey said they are either using or expect to use either a vendor solution or a hybrid of internally driven and vendor-driven solutions to inject AI into their DevOps platform.