AI Foundations for IBM i

At the outset of the year, we highlighted the potential for emerging AI-based capabilities for IBM i as a key trend for 2024. At the time, our experience suggested that manufacturing and distribution companies had not yet seriously engaged with AI solutions. Surveys suggested that AI was winning more headlines than IT departments, but we predicted that this trend would shift in 2024 — not due to the sweeping adoption of dramatic AI technologies, but the targeted use of AI-based tools for tasks like data analysis, repetitive task automation, and predictive maintenance.

Now, the latest announcements from IBM have confirmed that 2024 will be a big year for AI on IBM i systems. 

IBM Announces a Framework for AI on the IBM i Platform

During the keynote presentation from the POWERUp 2024 conference in May, IBM i CTO Steve Will announced that IBM planned to incorporate AI with the platform via three distinct avenues:

1. Db2 Data Analytics for connecting business transactions with AI-powered analytics, including trend analysis and anomaly detection, 
2. Operations, including active monitoring tools and self-healing capabilities designed to improve Power system performance.
3. Tools for supporting developers, including an AI assistant for coding creation and interpretation.

For the coding assistant, IBM plans base features including the ability to:

• Help programmers work with existing RPG code, including examining and explaining code.
• Generate modern free-format ILE RPG code based on a description.
• Write test programs for RPG.

Notably, the first function planned for the IBM i coding assistant will be to explain the purpose of a section of RPG code. We see this as a sign that IBM is not just jumping on the AI bandwagon, but thinking carefully about how the technology can drive value in targeted, relatively accessible applications that make sense for the prototypical IBM i business.

According to reporting from IT Jungle, Will noted in the keynote that “this is one of the issues we tend to find with new programmers going into old code…it doesn’t matter how well somebody thinks they documented it — if it was written in RPG 2, it looks really hard to understand to a brand new programmer. And so if you could have a tool that would do that, that would be very, very beneficial.” IBM plans for the RPG code assistant to:

1. Help early career and/or non-RPG programmers learn the language.
2. Provide support for current developers by assisting with modernizing code.

Evidence from IBM Software suggests that generative AI coding tools are capable of driving 30-40% productivity gains for developers, and we are excited to explore the potential of this new technology. 

While we expect additional announcements on these capabilities in the coming months, this framework will provide a helpful strategic guidepost as IBM i businesses begin evaluating opportunities for early-stage AI implementations that can drive value without large-scale investment or complex development work. 

Thoughts on the Future of AI on IBM i

TechChannel reports that Will projects “over the course of the next several months, couple years, we will have a tool that was specifically trained to help this community solve one of the biggest problems that we face, which is continuing support and enhancement of existing code that we have. We’ll start with RPG, and we’ll see where it goes from there.”

Moving forward, we see capabilities like AI coding assistants as not only potentially valuable tools but useful models for businesses seeking to dip their toes into the AI waters. As organizations begin to evaluate AI applications, it will be important to recognize that implementing select AI tools will not require redesigning IT systems from the ground up. In fact, as a tool like an RPG-coding assistant demonstrates, AI even has the potential to help older applications continue to perform their best.

Whether or not an RPG coding assistant is the right fit for your business needs, the following framework should help approach early-stage AI projects strategically. IBM’s planned tools appear promising, but solutions need not be IBM-specific to be integrated with existing systems.

1. Start by identifying relatively simple, repetitive tasks that waste time for human personnel and may be readily automated. Even a relatively simple document or invoice processing tool can potentially save hundreds of hours per year, reduce errors, and free up human personnel.
2. Leverage pilot projects to generate short-term ROI while building the long-term foundation for an organizational culture and skillset that are primed for more widespread AI adoption.
3. Identify opportunities to make use of employees freed up by AI tools to contribute to higher-value work.

Our team at Procern is looking forward to working with our clients to identify and drive value through AI-powered capabilities. ProCern can provide our clients with end-to-end guidance for deploying AI solutions that are targeted to solve specific business problems.

Learn more about ProCern’s end-to-end support for value-driven AI capabilities for fraud provention, quality control, process automation, and more.