Testing and the quality assurance field in general have evolved a lot over the years. Where are we today? For example with continuous testing in a DevOps environment. And where are we heading? I’d guess towards using AI for quality forecasting.

IT is never a goal in itself, it is meant to support “the business”. The business pursues value. To achieve that business value they need some IT. But how will the business people establish their confidence that the pursued value will indeed be achievable?

That’s where testing comes in. Based on the objectives we define indicators. Measuring these indicators by testing, will supply information which is used to establish the confidence level.

In todays DevOps world selecting the right indicators is paramount, because a large share of testing will be automatically executed in the continuous delivery pipeline, so wrong indicators would automatically generate wrong information.

When we have the right indicators will we have the right information in time? Testing gives information before the system goes live. But in the DevOps culture we don’t always test all important things before going live, instead we monitor the live situation. Thus there is a lot of information about the system behavior from both testing and monitoring. And this is the ideal starting point to take the leap towards quality forecasting.

Testing typically only gives information during the development of a system, and if there are anomalies, we’ll fix them before we go live.

Monitoring gives information while the system is live, and if there are incidents they can be fixed as soon as possible.

Now imagine that we could fix faults before any user observes a failure, wouldn’t that be great?
Using the data from testing and from monitoring, machine learning algorithms should be able to predict the trend in quality of the IT-system. This way AI will support quality forecasting. And using this forecast the people in the IT-team can indeed fix defects before they actually cause trouble.

Quality forecasting is the holy grail in quality and testing, and AI is our enabler. Let’s take the leap towards this future where finally testers are no longer the messengers of bad news (= defects) but the enablers of a stable quality level!!

Key Takeaways

• Continuous testing in DevOps is implemented in the pipeline, complemented by some exploratory testing
• Measuring business-oriented indicators on quality and risks supplies information that stakeholders use to establish their confidence
• Based on information from testing in the test environment and monitoring the live environment, AI will predict the future quality and based on this quality forecast defects can already be fixed before any user observes them