AI lead Software Engineering: Sopra Steria Ecosystem Offerings

Apart from internally developed IP’s given in the previous edition we also have an ecosystem of mature market leading companies, start-ups as well as labs and universities to build competency in their offerings and use them to help our clients. Following are some of the offerings from the ecosystem:

Axway Decision Insights (part of Sopra Steria group) is an industry leading, operational intelligence platform, which combines business activity monitoring with the business intelligence capabilities. It enables users to do on-demand simulation of a past situation with exactly the same information as when they occurred live and answer questions like what happened and when and why certain actions were taken. This bi-temporality is applied not only to external events but also to Decision insight’s internal objects such as dashboards and models. This is to help analyse how a situation was perceived at different point of time based on what was known then through the dashboards. This in turn is helpful to carry out investigations for audit, traceability and risk assessments based on past and present data.

Axway Decision Insights has many success stories. Monoprix, a major European retailer has used it to track orders internally for its points of sale and monitor product availability more effectively. The operational Intelligence features brought by Axway Decision Insight technology enabled visibility and understanding of the past, present and future performance of the supply chain, and assisted Monoprix to have an end-to-end visibility into the order processing process and helping it identify potential problems within its supply chain.

Intelligent RPA

Intelligent RPA is about applying AI technologies such as NLP, ML, Deep Learning etc to RPA. This is helpful in many scenarios such as while automating a task involving human intervention in a business process, interpreting a document, resolving tickets or performing diagnostics e.g. for ensuring that all the deployed field cameras are in working condition and so on. Computer vision is another feature useful for AI powered document extraction, i.e. interpreting screen elements which are part of an image, pdf or a virtual desktop environment. As part of this feature RPA software take screenshots, parse them using OCR and apply deep learning to interpret them and use them in the process automation. 

Sopra Steria has deep working relationships with market leading intelligent RPA vendors such as UI Path, Automation Anywhere, Blue Prism, which can in turn help implement various AI enabled software engineering tasks. 

Diffblue uses AI to automatically write suites of unit regression tests reflecting the current behavior of the code. Once baselined, this regression pack can then be run for new commits to determine if something that has been previously working is broken or changed.  It can give alerts if there are unintended changes in the code behavior so that developers can fix them. Post a successful release, a new test baseline can be created by a click of the button to keep the regression pack always up-to-date without manually maintaining them. This saves the organization from both manually creating and maintaining the unit regression tests resulting in cost savings, lower elapsed time during the development, testing and deployment activities. It also helps releasing the team’s bandwidth for more valuable and creative aspects of their work. Diffblue also helps increase the test coverage by automatically covering all the parts of the code thereby improving the quality of testing.

Repairnator. The Eclipse Repairnator project comes from the research activities by the University of Lille, Inria, University of Valenciennes, KTH Royal Institute of Technology. It is an open source project in the domain of intelligent continuous integration (AI for CI). At a high level, Repairnator is a bot for automatic program repair and all types for repair are considered such as test failures repairs, compilation errors repairs, static warnings repair, crash repair etc. Repairnator bot continuously monitors for the build failures on the Continuous Integration (CI) platform. Once a build failure occurs it collects and analysis the CI build logs, attempts to locally reproduce the build failure that has happened on the CI platform and then run different program repair prototypes built via academic research. When a patch is found, its validated and then submitted to the original developer as a proposed code modification by an external contributor. The original developer can then follow the usual process of code review and merge. 

Repairnator focuses on repairing Java software built with Maven. The fixes are generated by Nopol, Astor and NpeFix algorithms. Repairnator is an open source project which has already given fixes to open sources projects on GitHub which are accepted by original developers.

Read more

AI led Software Engineering: Sopra Steria Offerings