Amazon Q Developer guide
AWS is releasing a preview of chat customisation today, along with the general availability of the customisation feature for inline code completion in Amazon Q Developer (in your IDE). Now, in the IDE code editor and chat, you can set up Amazon Q to recommend specific code from private code repositories.
A coding companion with artificial intelligence (AI) is Amazon Q Developer. By providing code recommendations in integrated development environments (IDEs) based on pre-existing comments and code, it aids software engineers in speeding up the development of applications. Large language models (LLMs) that have been trained on billions of lines of code from open source and Amazon sources are used in the background by Amazon Q.
You can download the extension for JetBrains, Visual Studio Code, and Visual Studio (preview) and find Amazon Q in your IDE. You can write complete functions from a remark you enter, or it can suggest code as you type in the IDE text editor. Additionally, Q Developer allows you to speak with it and request code generation for particular jobs or explanations of code snippets from code bases you’re discovering.
Developers can now get even more pertinent code recommendations based on internal libraries, APIs, packages, classes, and methods of their company thanks to the new customisation feature.
Let’s take an example where a developer employed by a financial organisation is assigned to design a function that calculates a customer’s overall portfolio worth. Amazon Q will now recommend code to create the function based on examples it learnt from your company’s proprietary code base. The developer can now explain the intent in a comment or type a function name like computePortfolioValue(customerId: String).
In the conversation, the developer can also pose inquiries regarding the code used by their company. Assume for the purposes of this example that the developer in question is a new member of the team and is unable to obtain a customer ID. He is able to ask in simple English in the chat window: “How do a connect to the database in order to get the customerId for a particular customer?” Q&A on Amazon could respond: A discovered a function that uses the database connection XYZ to retrieve customerId based on the customer’s first and last name.
Using your internal git repositories (like GitHub, GitLab, or BitBucket) or an Amazon Simple Storage Service (Amazon S3) bucket, you as an administrator build customisations. It assists Amazon Q in comprehending the purpose, identifying the public and internal APIs that are most appropriate for the job, and producing code recommendations.
With its high data privacy and security, Amazon Q customisation capabilities lives up to your expectations from AWS. Your company retains private access to the code base that you share with Amazon Q. Their foundation model is not trained using it. After customisations are implemented, your organization’s developers can only access the inference endpoint in private. You won’t see recommendations in the developer IDE of another company based on your code. You control which developers have access to each specific customisation, and you can monitor metrics to assess how well your deployed customisations are working.
AWS capacity to customise Amazon Q was developed using cutting-edge technology methods including Retrieval Augmented Generation (RAG). Further information on the science underlying the Amazon Q customisation feature may be found in this incredibly thorough blog post.
Two new features have been added since AWS released the preview on October 17 of last year: updating a customisation and customising the IDE chat.
You want Amazon Q to consistently recommend the most recent code snippets because your organization’s code base is continuously changing. The AWS Management Console now allows an Amazon Q administrator to initiate an update procedure in just one step. In order to guarantee that developers always receive extremely accurate code suggestions, administrators can plan regular updates based on the most recent contributions on code repositories.
Developers in your company can select a section of code in their IDE and send it to the chat to get an explanation of what the selected code does using the new chat customisation feature (in preview). Additionally, developers can pose general queries about the code base of their company, such as “How do a connect to the database to retrieve customerId for a specific customer?”
Amazon Q For Developer
AWS choose to highlight the recently added customisation update feature that is currently widely accessible in this demo.
You go to the Customisations part of the Amazon Q console website in order to update an existing customisation. You choose which customisation you wish to change. Then you choose to Create new version under Actions.
Depending on how much code needs to be ingested, creating a new version of the customisation could take some time. A new version is available under the Versions page once it’s ready. You can choose whether to activate the new version so that your developers can use it by comparing its Evaluation score with that of the previous versions. Return to an older version anytime.
You monitor User activity on the Dashboard page. You’re able to keep track of many metrics, such as the number of daily active users, lines of code written, security scans conducted, and so forth. If, you are a former user of Amazon CodeWhisperer Professional, you may still see the CodeWhisperer name on some pages when using it today. It will gradually give way to the new moniker, Amazon Q Developer. More metrics are produced by Amazon Q and posted on Amazon CloudWatch.
Programming languages that are supported
At the moment, codebases with Python, JavaScript, TypeScript, and Java written in them can alter Amazon Q suggestions. When building the customisation or making customised recommendations in the IDE, files written in other languages supported by Amazon Q such as C#, Go, Rust, PHP, Ruby, Kotlin, C, C++, Shell scripting, SQL, and Scala will not be used.
Amazon Q developer pricing
AWS Region agnostic, developers can access Amazon Q globally. At the moment, Amazon Q is hosted in the US East (North Virginia). If your AWS IAM Identity Centre is located in a different Region, Amazon Q administrators have the ability to set up Amazon Q as an approved cross-region application.
The Amazon Q Developer Professional subscription includes free access to the Amazon Q customisation feature. Per AWS account, you can build and maintain up to eight customisations, two of which can be active at any given time.
