Oracle Code Assist, Java-Optimized, Now in Beta
LAS VEGAS – Here it comes again. Oracle Code Assist (OCA) is now in beta after an initial introduction in the spring.
At its user conference here, Oracle CloudWorld, the company also announced some incremental enhancements for Oracle Cloud Infrastructure (OCI) and OCI Kubernetes Engine (OKE) to enhance the deployment and management of AI workloads.
Like most tools in the space, Oracle says OCA helps coders code faster. But it arrives with its share of tools for developers. Java-focused in many respects, it has a potential fit for the enterprise developer.
Developers may use OCA with JetBrains IntelliJ IDEA or MicrosoftVisual Studio Code as a plugin. It is cited as optimized for NetSuite’s SuiteScript and SQL and used on OCI. OCA also supports JavaScript, Python, SuiteScript, Rust, Ruby, Go, C# and C.
What’s new? Oracle cites Java optimization as a way to build Java applications and update legacy software. It also cites application resilience, performance and security. Core features, announced in May, include:
- Code generation.
- Test case generation.
- Documentation creation.
- Automatic language upgrades.
- Code analysis for optimization.
- Code explanation for new talent.
In its announcement, the company did not reveal what it used to train OCA. But Oracle is relying on its deep roots in the Java community to provide automated recommendations in OCA.
What other information is being used to train the LLM? In the overall market, we see tools from providers such as Codium and Tabnine being trained on publicly available information. More should be announced this week at Oracle CloudWorld.
OCA will likely be used alongside other AI code assistant tools in an integrated developer environment (IDE). That contrasts tools used as stand-alone services accessed through a web browser. Additionally, it should be noted that Visual Studio and JetBrains have their plugins.
We’ll know more this week, but the big questions for Oracle include how OCA will be used for writing code, code review, debugging and help with specific coding challenges.
Oracle Kubernetes Engine Updates
Oracle continues to incrementally add features to its Kubernetes platform, OKE. In particular, it now has Ubuntu support and can deploy a GPU-based workload as an Ubuntu image.
This provides several advantages. Ubuntu repositories include NVIDIA drivers, allowing for native GPU support. It also offers the NVIDIA CUDA library to provide a way for accelerated GPU computing.
Ubuntu also offers a relatively efficient way to develop and deploy from a laptop or workstation to a cloud-based environment where Ubuntu is running. Cloud services widely use Ubuntu, which has long-term stability and is recognized by the AI and data science community.
Oracle also cites enhancements, such as the integration of Oracle Cloud Guard to extend monitoring capabilities for containers. The software uses machine learning models to find patterns that help with threat detection. It also includes OCI Logging Analytics integrations and node health checks and cycling.
AI Surge Continues
The story from Oracle CloudWorld encompasses as much as you’d expect from one of the world’s largest software companies. And much of it, like many other large software companies, includes lots of news about partnerships with other big software companies, including Oracle Chairman and CTO Larry Ellison talking on stage at Oracle CloudWorld with Amazon Web Services CEO Matt Garman.
But at the center of it all is the continued surge of AI into almost all of Oracle’s news. A code-assist tool and Kubernetes enhancements are just parts of the story.
Still, we are deep into the unknown, with survey research showing that developers express interest in using AI coding tools, but the types of products need to be more understood. AI does not vex developers. More so, they express concerns about the impacts on their salaries.
Lawrence Hecht contributed to this article.
