Every Novel Heat customer has the option to submit data to a global code repository. If you choose to do so, your source code will never leave your organization’s hard drives. Rather, it’s first abstracted and de-identified into patterns of behavior.
Each time your code is refactored, a senior member of your team tags it as such. This conveys to Novel Heat your team’s preference on how to write that piece of code correctly. Our solution monitors these patterns of behavior, noting how code is being rewritten.
Through machine learning, Novel Heat recognizes patterns in how code is being changed — both in your organization and in others around the globe. Whenever it identifies a weakness, such as a God object, Novel Heat suggests how to convert the devilish code into heavenly code. Getting this feedback helps every developer on your team become the best possible developer — and build the best possible software.
Novel Heat recognizes well-known anti-patterns, such as God objects, feature envy and dependency hell. Through machine learning, it identifies best practices that are used on your project along with architecture and behaviors that are considered bad practices. Our solution will make recommendations based on knowledge captured from teams worldwide, but still allow your teams to keep their own individuality. After all, what constitutes “best practices” is in the eye of the beholder.
