The client is a corporation operating in the field of designing, manufacturing, selling, and serving innovative analytical instruments for scientists that help them reach goals, increase productivity, and earn laboratory-based organizations a higher ROI in research, development, and quality control.
The customer sought to ensure high quality of the product by implementing an automated regression testing solution that was to increase coverage and shorten test cycles.
It was also required to automate software deployment and load testing. After evaluating other players in the market, the client chose to work with a1qa because of the company’s expertise in test automation, high-quality solutions, flexibility to work with third-party technology providers, and responsiveness.
In the scope of regression testing automation, building and maintaining a whole test infrastructure all by one team was challenging. To address this problem, the a1qa engineers selected Docker as an effective tool to set up and maintain servers for test automation.
The client was seeking test automation solution that would help ensure high quality of the company’s product – the solution that encompasses data acquisition, processing, visualization, reporting, and configurable compliance tools within a networked laboratory environment.
To address the challenge, the team of 10 software testing engineers got on board. To start with, they thoroughly investigated the solution under test and came up with a set of automated tests as a PoC, optimized the software test architecture.
The customer assessed the quality of the developed tests, greatly appreciated the a1qa personal approach and commitment, so the development of the automation solution was handed over to the QA engineers.
The a1qa team was provided with the automation solution developed by the client’s team years ago. Having reviewed it, the engineers came to the conclusion that it would take a lot of time to polish and finish it off. Therefore, it was decided to start developing a1qa’s solution from scratch without making use of the legacy one.
It’s important to point out that all test scenarios were developed by the a1qa test designers. They also identified and described the applied test techniques, tools, defined and maintained test automation architecture, verified the test approach. The documentation was later transferred to the client.
To estimate the results of the automation project, the client calculated the number of manual testing efforts that would have taken to do all the testing that was automated. These tests were running daily on multiple release branches.
Giving a total of 403 MTE hours per day or just over 300,000 MTE hours per year (3 branches * 403 hours * 5 days * 52 weeks), or the equivalent of 156 manual testing engineers working 8 hours a day, 48 weeks a year:
Besides automating laborious regression testing, the a1qa engineers developed a solution that allowed for automated comparison of two files.
One of the software functionality was the generation of HTML reports on different scientific works and then converting the report into the PDF format.
Previously, the client’s end-users had to make sure that the information had been converted correctly by hand, visually comparing all symbols, characters, and punctuation marks.
The automated solution released the scientists from this tedious work and reduced the chance of an error.
While the automated solution was under development, the QA managers from a1qa were engaged to rebuild and optimize the customer’s in-house QA processes. The main goal was to shift them from the traditional Waterfall model to Agile methods.
As a result, QA was shifted left and assured earlier defects detection. The QA engineers got an opportunity to provide rapid and continuous feedback on the testing results.
Agile sprints also benefited the company by delivering working software earlier, providing better estimates in less time, and allowing for any corrections instead of derailing the project.
The Agile methods implemented by a1qa were then successfully scaled up and employed in other client’s projects.
The client’s solution was made of several modules and their manual deployment took a lot of time. Therefore, the client asked the a1qa team to develop an automated solution to speed up the deployment of virtual machines and all the apps on them.
Once the deployment was over, load testing scripts were to be launched. To complete this task a dedicated team comprising of two performance testing engineers and a QA manager was assigned.
They developed the scripts for deployment of the required number of virtual machines in the AWS Elastic Beanstalk environment.
From this time on, the client’s systems were installed and connected to each other automatically. The load testing scripts were also launched without the direct involvement of the engineers.
The overall software development process has been orchestrated in compliance with the best Agile practices.
