The name of the base system (platform): | Amazon Web Services (AWS) |
Developers: | Amazon |
Branches: | Information technologies |
Technology: | Development tools of applications |
2020: Start of service
At the end of June, 2020 the Amazon company started the tool for CodeGuru application development. He uses algorithms of artificial intelligence which make to the programmer recommendations about improvement of quality of the code. Service works based on cloud infrastructure of Amazon Web Services (AWS).
Development teams check the code to estimate to his logician, the syntax and style before adding it to the existing code base of the application is a standard practice. But it is often difficult to find developers who could execute the overview and trace applications after deployment. Besides, there are no guarantees that developers will not pass errors which can lead to serious problems with performance.
CodeGuru solves this problem, being integrated with the existing development environments. This tool uses AI algorithms trained for more than 10,000 projects open source to estimate the code directly in the course of its writing. At detection of the error CodeGuru offers the legible comment which explains a key part of the problem, and offers possible methods of correction. The tool also finds inefficient and unproductive code lines which can lead to excess loading of the processor.
A system consists of two parts. CodeGuru Reviewer reveals a deviation from advanced methods of use API- the AWS interfaces and SDK and notes common problems which can lead to production errors. CodeGuru Profiler provides specific recommendations about such questions as use of inefficient libraries. It profiles the application and reveals problems with quality of the code (providing them together with detailed information on a delay and loading of the CPU).
Amazon claims that CodeGuru was used for optimization of 80,000 applications and allowed the company to save tens of millions of dollars. According to developers, some commands could reduce use of the processor by 325% and reduce costs for 39% of all in a year.[1]