Oracle is getting into the coding assistants game. The aptly named "Oracle Code Assist" is going to be optimized for Java, SQL, and Oracle Cloud-based applications.

As described in "Oracle Database 23ai - Feel the Difference", AI has become a top priority for Oracle which is looking to bake it into all of its products.

The start has been made with adding AI Vector Search to 23ai which makes it possible for LLMs to query private business data using a natural language to provide accurate and relevant results. This integration with LLMs such as Cohere and Llama that lets you ask questions using natural language, aims to make the wealth of information stored in your enterprise database approachable to anyone, as easy as chatting with ChatGPT. Therefore even those who do not understand SQL can easily and quickly query business data and generate reports.

The new endeavor is even more developer-driven, to aid them in their day to day coding tasks. It is inteneded that Oracle Code Assist will offer the following functionalities:

• Code Generation: Provide code suggestions based on learned context and patterns and generates entire class files, methods, or lines of code. Developers can then accept, modify, or reject the code snippet, reduce the time spent writing boilerplate code, and build applications faster.
  
  
• Code Annotation: Generate application programming interface (API) documentation or comments to describe code functionality. This will help developers with the application’s lifetime maintenance.
  
  
• Code Explainability: Provide a plain language summary of the code. This will help developers tasked with reviewing or refactoring existing code and saves time by providing precise and easy-to-understand context.
  
  
• Faster Code Reviews: Initiate a pull request to merge changes into a repository after writing code reviews and recommends code changes based on performance, quality, efficiency, and styling guidelines. This will help developers deliver higher-quality code.
  
  
• Test Coverage Generation: Create unit and functional tests based on the context of the code. This will help developers improve test quality with better test coverage spanning a greater variety of test inputs.

If you watch the demo presentation of the assistant you're going to be impressed. For instance, it can generate a Spring Boot app together with WebSocket support, generate the relevant unit tests and write the documentation too.

A feature that will be very handy is that developers will be able to use Oracle Code Assist to quickly upgrade entire Java applications to newer versions. This ability to rapidly upgrade Java using a coding assistant significantly increases the ability of enterprises to improve application resilience and performance, incorporate the latest JDK features and bolster their security posture.

I have a few pre-version 3 Spring Boot applications as well some Java 8 apps that I would like to upgrade and auto-upgrading would come very handy for sure.

While there's an explosion of code language models, as we've described in "IBM Launches The Granite Code LLM Series", it is not yet revealed which underlying model Code Assist is going to plug into. The privacy guardrails are also not disclosed, something pretty important when applying the model on your own code. And finally pricing is still TBA.

In any case, the landscape is heating up reaching boiling point, as all of the big players are pumping up their products with AI capabilities.

Oracle's Coding Assistant As discussed in Software Engineering Will Never Die, by having tools that are capable of writing large parts of applications, it increasingly looks like that the programmer's job will shapeshift into that of a plumber who pieces the individual, auto generated or tweaked parts together into one final product, whilst keeping in line with the business rules.

More Information

Oracle Code Assist

Related Articles

Oracle Database 23ai - Feel the Difference

IBM Launches The Granite Code LLM Series

Why Software Engineering Will Never Die

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.

Comments

Make a Comment or View Existing Comments Using Disqus

or email your comment to: comments@i-programmer.info