IBM introduces the watsonx Code Assistant for i: an AI tool that supports developers in modernizing IBM i applications, focusing on RPG code.
IBM announces the IBM watsonx Code Assistant for i. This is an AI-based programming assistant designed to help organizations modernize IBM i applications. The tool is currently in private preview and primarily offers support in explaining RPG code.
Help for Scarce Developers
The IBM watsonx Code Assistant for i aims to simplify the development and maintenance of IBM i applications. IBM i is still widely used globally for business-critical workloads, according to IBM itself, due to its stability and low ownership costs. Many of these applications are written in RPG: an older programming language for which developers are increasingly difficult to find.
read also
IBM Launches New AI Mainframe: Smarter, Faster, and More Energy-efficient
The AI assistant utilizes generative AI and is integrated into development environments such as Visual Studio Code. In the current testing phase, the tool provides contextual explanations for RPG code. Later versions will include features like automatic code generation, unit test creation, and transformation of existing code.
Granite and IP Protection
The underlying AI models are based on IBM’s own Granite code platform. It is specifically trained with RPG code and IBM i resources. According to IBM, this should ensure more relevant support in the development cycle. The assistant also provides support for DevOps processes and IBM i modernization tools.
Companies can choose how they want to deploy the IBM watsonx Code Assistant for i: in the cloud, on-premises, or in a hybrid environment. This allows for consideration of data governance and regulations. IBM also emphasizes that the Granite models come with IP indemnification. In other words, users are protected against infringements on their intellectual property when using the AI model.
With the new tool, IBM targets developers with varying levels of experience. As the initial focus is on context and explanation, the company clearly aims to assist less experienced RPG programmers as well. In this way, AI should somewhat compensate for the lack of knowledge and experience in the market.