Amazon Unveils ML-Based Code Generator ‘CodeWhisperer’


The service can generate more than ten lines of code at once and bases its recommendations on a variety of contextual cues

Amazon recently announced the launch of Amazon CodeWhisperer, an AI-powered coding aid service by Amazon Web Services in addition to its cloud offering. To assist programmers in writing code more quickly and creating training datasets for their artificial intelligence projects, the service presently in preview, was created. Amazon CodeWhisperer uses AI to generate software code that would otherwise need to be written manually by developers, saving them time in the process. As a result, software projects can be finished more quickly.

CodeWhisperer was brought into the public eye at re:MARS 2022. Billions of lines of code from open-source projects, internal code repositories, and other sources have been used to train CodeWhisperer. The service can generate more than ten lines of code at once. CodeWhisperer bases recommendations on a variety of contextual cues. The cursor position in the source code, the code that comes before the cursor, comments, and code in other related files are just a few examples. The AWS IDE Toolkit is a tool developers can use to access Code Whisperer. It is compatible with each developer’s integrated development environment (IDE). In addition to numerous programming languages, including Java, JavaScript, and Python, CodeWhisperer supports many well-liked IDEs.

One of the use cases for CodeWhisperer is the creation of code snippets from text inputted in natural language by developers into source files. CodeWhisperer may automatically locate the required technologies and include them into the code, even if performing a task necessitates utilising other technologies like a cloud service or library. Making it easier for developers to use the cloud platform offered by AWS is another application for the service. Tasks like building buckets or data storage repositories in AWS’s well-liked Amazon S3 storage service are more straightforward with CodeWhisperer. It also integrates with other well-known services like AWS Lambda and Amazon EC2.

Along with CodeWhisperer, an update to SageMaker Ground Truth, an existing AWS service, was also made available at re:MARS 2022. The solution aids software development teams in collecting training data for their AI initiatives. The process of creating an AI application is lengthy and comprises many steps. Preparing a training dataset, which can take months because the data frequently needs to be created manually, is one of the most time-consuming tasks. Annotating these training datasets is a whole other step. According to AWS, the current update to SageMaker Ground Truth will ease this task.

SageMaker Ground Truth will have capabilities for creating synthetic data as part of the release. To hasten the development of AI, businesses can now request that AWS construct a synthetic training dataset on their behalf. Together with CodeWhisperer, the new synthetic data production capabilities will broaden AWS’ already substantial AI portfolio. Numerous services that facilitate neural network construction are currently available from the cloud juggernaut. Through services like Amazon Personalize, which can be used to generate product recommendations automatically, it also offers pre-packaged machine learning capabilities. The latest improvements will significantly increase what AWS already provides.