Competency

Automated testing

Test automation at all levels of testing, from unit to end-to-end testing, is essential in order to deliver new functionality and bug fixes continuously, while lowering the risk of releasing expensive defects into production.

Both functional and non-functional requirements must be covered by automated tests in an efficient and maintainable manner. We have the experts that can show you how, based on solid, professional competency and experience from a number of large and mission critical projects within both technical and administrative domains.

Model-based testing

Using model-based test techniques we can automate the test automation! The result is higher test coverage more quickly, simpler maintenance of test code, as well as more correct and precise requirements, and more effective collaboration between business analysts and developers who together build a shared understanding of the system under test.

Models can in many cases be quite simple, while it’s also possible to model more complex solutions and thereby attaining test coverage that is impossible to reach using traditional methods. The benefits of using model-based testing are usually greater the more complex the system under test is!

Several of our employees have long formal and practical experience in this area and can help you get started! Model-based testing is not a single technique, but can range from very simple models and associated test case design, to complex models of for instance distributed real-time systems, where automated code generation and verification of attributes such as safety, latency, security etc. is essential. One example of a relatively simple but very powerful model-based test tool is our own open-source tool and software-as-a-service: ecFeed!

Smart testing with machine learning

With our in-depth knowledge of machine learning, we can deliver a range of solutions to streamline and improve your testing.

Using machine learning we can analyze large amounts of data and develop solutions for smarter testing! That includes error prediction for more focused and risk-based testing, test selection for regression testing and retesting, generation of test oracles where you analyze test results and return an automated pass/fail, automated security testing that searches for security holes, generation of large volumes of rich synthetic test data, grouping and analysis of test results, and a lot more!

Several of our solutions have been publicized in international conferences and journals.

Test data generation

An important prerequisite for successful testing in general, and automated testing in particular, is that you can make high-quality, representative and complex test data that cover the necessary variations and volumes needed for good test coverage. This applies to functional and non-functional testing alike.

Several of our employees have expertise in this area and have worked with both generation of synthetic test data based on machine learning techniques, domain-specific languages, rule-based anonymization techniques and much more.

We also have an ongoing industrial PhD project in test data generation in collaboration with Norwegian public sector departments.