AI&ML-enabled DecSecQAAIOps

For the last few weeks, I have been wondering if we are doing QA automation and DevSecOps right? Is it futuristic? Are we leveraging the new age technology capabilities in our product development life cycle? I am sure my colleagues who are leading new generation SaaS products are wondering too.

AI&ML-enabled SaaS products: These products are not only dependent on data but also sensitive to data, both historic as well as to the live input data streams. System behavior in production changes as the data catachrestic changes. Now operation concerns are not limited to system crashing and abrupt behaviors but also about data and model output quality which adds to the operation's system reliability engineering complexity.

DevOps to DecQAOps to DevSecOps to DecSecQAAIOps:We have experienced the DevOps to DecQAOps to DevSecOps to DecSecQAAIOps journey where QA Automation and Security tools were integrated into the CICD pipeline to improve the overall quality and efficiency of the product development lifecycle. We are also observing a lot of automation and scripting work is now been done in operations by the system reliability engineering teams to improve system uptime, data, and model quality.

AI&ML-enabled DecSecQAAIOps: AI and machine learning in testing and its integration in the DevOps space is getting traction with new categories of testing tools. Some of these tools are classified under 1)RPA (robotic process automation), 2) NLP (natural language processing), 3) MBTA (Model-based test automation), 4) AT (autonomous testing). These tools are also categorized per their uses:

Differential testing also known as fuzzing is a popular software testing technique that attempts to detect bugs, by providing the same input to a series of similar applications. A few popular tools are Google OSS-Fuzz, Launchable, DiffBlue etc.

Visual AI testing addresses a user experience layer of testing and scales the validations and look and feel of a UI (user interface) across digital platforms (mobile and web mostly). A few popular tools are Applitools, Percy.

Declarative tools used AI&ML to enhance test automation productivity and stability. A few popular tools are UIPath, Automation Anywhere, Tricentis, Functionize

Self-update QA automation tools are AI&ML-based tools that update themselves as the product changes. They are mostly based on a record and playback mechanism with an AI&ML engine that updates that script as it detects UI changes. A few popular tools are Perfecto, Mabl

To close the DevOps look we need reporting and monitoring. ML in reporting helps sort through the data, slice and dice it, and in advanced cases, also automatically classify the root cause of failures and boost team productivity. A few popular tools are Perfecto, ReportPortal

To conclude: AI&ML-based QA tools are augmenting and complimenting current QA automation processes in the CICD pipeline. Its use in operation system reliability engineering is also prevalent in SaaS product organizations.