- Home >
- Case studies >
- Booking and distribution system
QA for booking management software
Award-winning booking system for tours and attractions.
90%
Test coverage
3.2k+
Automation test scripts created
1-2h
Time of regression testing run
160
Amount of devices used
70%
of blocker/critical/
major bugs found
5-10
Releases to Prod per 5 working days
2.4k+
Reported blocker/critical/major/minor bugs
Thanks to the use of AI, it was possible to modernize the set of old Selenide + Cucumber tests with the aim of obtaining features of modern frameworks.
Namely, the goal was to organize an internal re-run of failed scenarios inside the job and stop the execution of tests if a large number of tests failed within one job on the dev branch. Using ChatGPT, a complex Bash code was written that performs the functions described above on the GoCD pipeline.
The result of this is a reduction in the execution of the automation build before release for the main application from 1 hour to about 20 minutes, and also allowed saving time (that is, money) when using remote agents.
About project
Xola is a platform designed to simplify the operation of tour and event companies. It optimizes a variety of tasks including reservation management, payment processing, schedule tracking, inventory management, and customer relationship management.
Xola’s primary application enables in-person reservations, booking management, tour operations, and marketing promotions. Additionally, there's an online booking app that's highly adaptable and can be seamlessly integrated into a seller's website to check available tours with various customization options.
Before DeviQA
A product was unstable and loaded with bugs
No automated tests
No regression testing
No CI environment
No parallel threads
With DeviQA
Fast and stable automated regression testing
Ability to make releases 1-2 times per day
75% of bugs are detected at the development stage
3200+ automation scripts created
8300+ checks added
2 hours to run automated regression testing
Best practices in automation testing:
Page object pattern
Independent tests
API data creation
Clear HTML reports
Regression testing for 6 apps was implemented using advanced automation testing technologies
6 CI pipelines were configured in collaboration with DevOps engineers
70+ parallel threads
Our contribution
Team
6 full-stack QA engineers
Project length
Since 2014
Technologies and tools
Jira
Confluence
Java
Selenide
Selenium
Cucumber
REST Assured
Gauge
GoCD
Postman
FullStory
Our engagement
Xola's partnership with DeviQA has significantly improved the stability and robustness of its product.
We've improved the description of tasks to make them clearer for a team. Bug reports now contain all necessary details and attachments as well.
We've written a lot of new test cases for sub-apps and created a comprehensive suite of over 3200 automation scripts with more than 8300 checks. This enables us to complete regression testing in just 2 hours, allowing Xola to release updates 1-2 times a day.
We follow best practices in automated testing like the Page Object pattern, test independence, API data management, and informative HTML reports. We've automated regression testing for 6 different apps, set up 6 CI pipelines with the help of a DevOps engineer, and improved efficiency with over 70 parallel threads.
Our meticulous bug-tracking process has identified over 2400 issues, with 70% of them categorized as blocker, critical, or major.
Our dedicated team of 6 Full-Stack QA Engineers utilizes advanced technologies and tools such as Jira, Confluence, Java, Selenide, Selenium, Cucumber, REST Assured, Gauge, GoCD, Postman and FullStory, ensuring the success of Xola's projects through thorough quality assurance.
Thanks to the use of AI, it was possible to modernize the set of old Selenide + Cucumber tests with the aim of obtaining features of modern frameworks. Namely, the goal was to organize an internal re-run of failed scenarios inside the job and stop the execution of tests if a large number of tests failed within one job on the dev branch. Using ChatGPT, a complex Bash code was written that performs the functions described above on the GoCD pipeline. The result of this is a reduction in the execution of the automation build before release for the main application from 1 hour to about 20 minutes, and also allowed saving time (that is, money) when using remote agents.
Services provided
Web automation testing
Web testing
Mobile testing
DB testing
Dedicated QA team