Today’s developers are constantly expected to do more with less. With the advent of new tools and as businesses look to build new revenue streams and innovations, developers are expected to write more code, at a higher velocity, without sacrificing quality. Unfortunately, writing high-quality code requires significant time committed to testing – time developers don’t have. To make matters worse, if a code defect somehow slips past testing and into production, it can be extremely expensive and time-consuming to fix down the road.

Testing code is not a fun process. Traditionally, developers finish writing blocks of code, check it into a version control system such as GitHub, and enter a rigorous testing phase where they run their work through numerous trial scenarios in an effort to break it. When bugs are detected, developers must go back into the code, try to fix the defect, and restart the testing process all over again. It’s inefficient, time consuming, and expensive.

Altran, the world leader in engineering and R&D services and a Microsoft Gold Partner, looked at this process and knew that there had to be a better way to test code—a method that allowed developers to identify bugs as they worked rather than waiting for the testing period.

Introducing the Code Defect AI solution

Altran’s new Code Defect AI solution allows developers to identify bugs earlier in the development lifecycle than ever before. As developers write code, this solution uses Azure AI and machine learning (ML) technology to check for defects, predicting where the code is most likely to break during testing. It does this by comparing the source code with decades of historical data, noting code patterns that have caused defects in the past. This helps developers take proactive measures to fix bugs before entering the testing phase, improving efficiency and helping them reach production faster.

This is, of course, a high-level overview of the solution. For more details on how this solution works, visit the Microsoft AI Lab site here.

Partnering with Microsoft

Altran has worked with Microsoft to develop products for over a decade, but that isn’t the only reason they decided to team up on this solution.

First, Microsoft’s immense pool of developers made them the ideal partner for building and testing this developer-centric solution.

Second, the pre-built power of Azure AI lent itself perfectly to Altran’s needs, helping them train their ML models quickly without having to start from scratch.

Third, Altran knew Microsoft could provide the troubleshooting support required to deploy the solution on GitHub.

Finally, Altran and Microsoft share a common goal: enabling the best experience possible for developers. Both organizations believe that by helping people work smarter, they can empower the world to achieve more.

Looking ahead

The Code Defect AI solution is already ground-breaking and has been rolled out to a few Tier 1, UK-based communication service providers, but Altran and Microsoft are just getting started. The next step is to couple the solution’s defect-prediction abilities with the ability to prescribe specific tests. This will tell developers exactly what tests to run to find their bugs rather than just predicting where defects are most likely. This vision is currently a work-in-progress, but when it’s ready in a few months, Altran and Microsoft plan to make it immediately available for internal users. After a short refining period, it will be available to customers worldwide, so stay tuned.