An end-to-end solution from testing to adoption
Sopra Steria has supported Credem through its journey to transform its processes, beginning with experimentation, including co-design, followed by the gradual adoption of generative AI technologies – including Azure OpenAI for semantic analysis and generation – that changed how it works.
The project began with a Proof of Concept focused on functional testing, with the aim of rapidly demonstrating the value of generative AI in the teams’ day-to-day activities and offering natural, non-intrusive support for their work. AI was used to analyse specifications and automatically propose verification scenarios, allowing the creation of complete and consistent test books.
As part of the proof of concept, Sopra Steria and Credem also trialled further automation, using browser automation tools such as Playwright and Selenium to generate and execute test scripts. This remains a possible future expansion of the programme.
Alongside this, Sopra Steria is supporting Credem in how its rethinks its processes, introducing AI as ongoing support in project management and process analysis activities, assisting with the creation of documentation, alignment with compliance requirements, and the simulation of possible evolution scenarios.
Although the first deliverable has now been delivered incremental improvements will continue to take place. Meanwhile, other projects that look at project management and knowledge graph integration are in early stages of development.
Testing effort reduced while time to test accelerated
The introduction of generative AI has reduced the effort associated with manual testing activities by nearly a third (30%), while the automation of specification analysis and test-case generation has led to a more than threefold acceleration in the creation of test books, in the areas where it was applied.
The level of functional coverage also increased substantially as AI was able to identify additional verification scenarios that might not emerge from entirely manual processes.
The standardisation introduced by AI has also improved the quality of project documentation with outputs that are more consistent, coherent, and easier to read.
Most importantly, it has achieved what it was meant to do – aiding Credem’s employees to improve how they work, enabling teams to switch their efforts from repetitive operations to higher-value activities that allow them to better respond to change. The project confirms that artificial intelligence is already a lever capable of making IT processes more scalable, intelligent, and reliable, while also time strengthening governance and the role of people throughout the application development lifecycle.