Combining quality with speed is a very desirable premise while developing software. However, this is not always easy to achieve. Nowadays, digital transformation and the need to adopt an innovative and customized business approach make companies develop software tools focused on the client, instead of aiming only at internal users. In this context, avoiding errors in apps and websites has become a crucial matter. But how can this be accomplished? By strengthening testing through testing automation development. Has your company already addressed this challenge?

It is known that qualified workforce is hard to find in the software industry. For example, a survey in Argentina showed that even though developers are the most required, testing specialists attract a significant portion of the searches (5 %). These profiles are not found in large numbers and their salaries are onerous. Over the last few years, a tester’s task has professionalized and currently implies complying with certain working standards and international certificates. In the national territory, there are only a few companies –such as Arbusta– which provide technology services specialized in testing. Due to this situation, many companies seek to run tests with fewer workers. How can this be achieved? Through testing automation.


As a general concept, automated tests are processes which validate whether a software works properly and complies with requirements before launching it into production. They can be applied to unit, regression and APIs testing. Their main advantage is they simplify the task of quality assurance teams. Test automation development provides a higher degree of accuracy, expands the possibility of obtaining reports, provides better coverage and efficiency, optimizes error detection (especially at early development stages) and increases reuse. Furthermore, it helps conducting faster verifications and avoiding human errors. Thus, it is not by chance that test automation deployment in websites and apps is becoming a trend. In terms of markets, this evolution makes more sense for those sectors which must provide an optimal client experience. This is why nowadays test automation is usually seen in banking, e-commerce and health services, for example.


Research shows that almost one fourth of IT budgets is already assigned to different artificial intelligence (AI) projects. Especially in the case of software testing, the use of automation tools based on AI makes tests stronger and more powerful (with a larger capacity for the detection of significant errors before running the tests), accelerates execution time and facilitates maintenance. Adding AI to testing automation is important, but not enough. Another key guideline to boost speed and productivity in software construction is to adopt an agile culture and a DevOps methodology. An efficient and automated continuous integration and continuous delivery (CI/CD) pipeline must also be structured –that is, a software development practice which enables receiving automatic notices upon changes and helps detecting errors in early stages. Through these processes, regression and performance testing can be automated and automatic controls of code quality, security, accessibility and usability can be included. 


Evidently, manual tests will not be completely replaced by testing automation; they will only be boosted and complemented by it. The human factor will always remain necessary, especially in those stages where in depth tests are required in order to ensure security and accuracy of all apps and technology based on data. And since AI starts to learn from the set of training data provided, such set must be carefully chosen, and this will require human involvement. Nevertheless, to a certain point, incorporating AI to testing shall enable non-technical staff to collaborate in the creation and execution of tests.

At a maintenance level –a stage which according to research usually takes up 30 % of testers’ time– tests based on AI replace static locators by dynamic ones. That is, an element is identified by several features on a page, so that said element can be located in case any feature changes. This brings more stability and helps saving a considerable amount of time.

AI also helps connecting production apps to the testing cycle, which enables the creation of tests based on real flow. AI is capable of observing and finding repeated steps and gather them in order to make reusable components. In addition, the more tests, the more the intelligence grows, optimizing the wait between steps and solving problems proactively. 

Using AI in test automation is a progressive trend which, apart from bringing efficiency, agility and productivity, allows to focus quality control manual efforts on those tasks with a higher aggregate value.

By Debora Slotnizky@deboraslot

> About Arbusta <

> What makes us special <