About the IBM I Coding Assistant. 

 Earlier this year, IBM announced a framework for AI on the IBM i platform, including:

• AI-powered analytics for DB2 databases.
• Active monitoring tools for operational support.
• An AI assistant for coding creation and interpretation.

We take a deeper look at IBM’s vision for AI in IBM i here. In this article, we focus specifically on additional details published by IBM on the upcoming coding assistant.

IBM i Coding Assistant: Key Capabilities

According to IBM’s official announcement and follow-up reporting, the upcoming RPG Code Assistant (RPGCA) is based on a large language model (LLM) specifically trained to understand RPG. It will eventually offer:

1. The ability to help programmers work with existing applications by analyzing and explaining the logic in existing RPG code.
2. Generative capabilities for creating RPG code based on a description, including converting older RPG formats to modern free-format ILE RPG.
3. The ability to write test programs for RPG.

Broader evidence from IBM suggests that AI-based coding tools are capable of driving substantial productivity gains for developers (up to 40%). Beyond these direct productivity gains, we think the greatest value of AI coding support for IBM i businesses will lie in its ability to address longstanding challenges for the IBM i ecosystem.

How Will AI Coding Fit into the IBM i Ecosystem?

IBM i continues to offer a state-of-the-art platform for a wide variety of manufacturing and distribution businesses. However, many organizations continue to rely on legacy applications that are not capable of tapping into the full potential of the platform. 

In this context, targeted application modernization and feature enhancements represent a high-value opportunity for many IBM i businesses. Efforts to update code, however, face common challenges related to the market for RPG expertise. 

• Knowledgeable resources with experience working with legacy applications for IBM i are already in short supply. With an impending retirement crisis, the demand for veteran RPG resources will only rise. RPG developers’ scarce time is often devoted to keeping
  business-critical applications up and running, with little time for modernization or enhancement work.
• Newer developers often have little experience working with RPG, and may find themselves challenged to understand older code written in an outdated format by a long-departed colleague.
• Relatively slow delivery times. Newer techniques like reusable objects and low-code solutions can dramatically accelerate workflows, but options capable of working with legacy RPG programs are limited.

How AI Tool (RPG code Assistant) will Help RPG Developers?

AI can directly address each of these challenges. By enhancing the productivity of developers, AI coding support can help experienced RPG resources generate as much value as possible. For developers without much prior exposure to RPG, AI support could make it dramatically easier to bridge the knowledge gap and make meaningful contributions.
For developers working to modernize legacy applications or develop new features, AI coding support has the potential to dramatically accelerate onboarding, execution, testing, and overall time to delivery.

How AI Coding Can Drive Additional Value from Existing Resources

AI coding assistance has the potential to not just accelerate workflows, but expand the capabilities of existing resources. While a select few developers can learn just about anything on the fly, the majority are most confident and effective within a select range of technologies. But with the right tools, a much broader set of programmers will be willing and able to contribute to a much broader range of projects.

For example, based on our experience supporting ERP software and other business-critical applications, general RPG programmers may lack the knowledge to implement new functionality via a browser-based web service. But these same programmers would be perfectly capable of working with an AI assistant that can help extend their RPG expertise knowledge to a range of front-end UI and feature enhancements.

These potential efficiencies will also help third-party service and support providers deliver additional value. For example, when ProCern takes on ERP support for a new client, we begin with an in-depth discovery process designed to help us learn the business from the ground up and the enterprise applications they depend on. AI tools for explaining existing code could substantially accelerate our timeline for onboarding new systems to support.

A Promising Use Case for Organizations New to AI

As we discussed earlier this year, relatively few manufacturing and distribution companies have begun to experiment with AI. We suggested that now is the time to start identifying early practical applications — use cases capable of driving short-term ROI while building organizational familiarity with AI tools. 

IBM i’s new coding assistant is a perfect example. Tapping into this new source of value doesn’t require businesses to give up on proven applications. In fact, it can directly support the enhancements that will keep these enterprise applications performing admirably for years to come.

The ProCern team is eager to collaborate with clients to unlock new value through AI-driven solutions. ProCern can offer detailed guidance for implementing AI technologies tailored to address specific business challenges. 

Explore ProCern’s end-to-end support for practical AI solutions for fraud prevention, quality control, process automation, and more.