Talsco Weekly: It's Time to Build Your First AI Agent

Welcome to another edition of Talsco Weekly

IBM i Brief: 💻 Power11 shows modest price/performance gains over Power10. 🔄 IBM wraps up 2025 with Power Systems updates.
AI: ✨ Writing an LLM agent is surprisingly simple and you should know how to do it.
Career: 💭 AI’s Promise and Peril.
ERP: 🔄 Infor unveils IBM i ERP roadmap at conference.
Hiring: 💼 Hire for curiosity, not your mirror image.
Modernization: 🔧 Legacy systems still dominate in 2025. 🔮 AI agents expose hidden gaps in legacy systems.
Open Source: Why Open-Source Software May or May Not be the Right Choice.
Security: 🔓 AI companies leaking secrets on GitHub. 🔖IBM i Security Pages.
Trends: 🔮 The Best-of-Breed Future: Composable Systems Over Monolithic Platforms.

IBM i Brief

💻 Power11 shows modest price/performance gains over Power10

IBM’s new Power11 entry and midrange servers deliver 10.5% and 5.9% better cost-per-CPW ratios respectively compared to Power10 predecessors when configured with eight cores running IBM i.

However, four years after Power10’s launch, the improvement falls short of historical Moore’s Law advancements. IBM i subscription costs range from $1,918 to $5,104 per core annually, with the lack of public pricing making system comparisons increasingly difficult for customers and partners alike.

🔄 IBM wraps up 2025 with Power Systems updates

IBM announced several end-of-year changes to its Power Systems platform, including a new AMD W6499 graphics card option for Power11 systems available December 12.

IBM also released updates to PowerVM hypervisor components and introduced IBM Concert Subscription for Power, an AIOps tool for Power infrastructure management. Meanwhile, IBM withdrew various hardware features including some Made in USA manufacturing options and multiple PCI-Express 4.0 Ethernet adapters, signaling ongoing platform rationalization.

AI

✨ Writing an LLM agent is surprisingly simple and you should know how to do it

There are big ideas in computing (and on the IBM i platform) that are easy to get your head around:

“S3 is to AWS what DB2 and single-level storage are to IBM i — a simple, universal storage primitive that the entire platform is built around.”

Yet, there are Other technologies, you need to get your feet on the pedals first. LLM agents are like that.

As we continue to cover, the IBM i platform is forever merging and integrating other technology stacks. This is certainly true of it’s integration with AI and LLMs.

For example, Project Bob

Here, Thomas Ptacek argues developers should build their own agents to truly understand the technology.

The basic code requires just 15 lines—an HTTP API call with a context array.

Adding tools like command-line functions is straightforward, and agents can autonomously determine how to use them.

Unlike complex AI services, individual developers can experiment with agent architectures, context engineering, and design patterns in their basement, making this an accessible frontier for exploration and innovation.

Career

💭 AI’s Promise and Peril

I’ve been wrestling with a paradox:

AI is simultaneously our greatest opportunity and our most significant disruption.

On one hand, AI offers organizations unprecedented capabilities to streamline operations and modernize legacy systems. For IBM i shops specifically, AI agents can transform how we interact with decades-old applications, making them more accessible and efficient.

But, I can’t ignore the broader view. AI will disrupt the workforce—that’s not speculation, it’s inevitable.

The question is: at what cost?

I sense a hesitation among organizations, maybe it is a strategic pause and more than likely, they don’t fully understand how to implement AI responsibly. See the AI Agents in Legacy Systems article below.

Or maybe unknowingly, they recognize the profound economic implications of workforce displacement and are reluctant to pull that trigger.

Organizations hesitate to deploy AI due to economic uncertainty, regulatory ambiguity, labor market tensions, social pressures, and questions about whether current AI capabilities justify wholesale transformation.

These macro forces create a perfect storm of hesitation—where the technology is ready, but the organizational, social, and economic ecosystems are still adapting.

We’re at an inflection point where technological capability outpaces our ethical frameworks. The challenge isn’t just can we implement AI—it’s should we, and how do we do so in ways that enhance rather than replace human potential?

ERP

🔄 Infor unveils IBM i ERP roadmap at conference At inPOWER 2025, Infor outlined plans for its IBM i-based ERP systems (XA, LX, System21), including 2026 support for IBM i 7.6 and new REST APIs. A key focus is migrating users to the Java-based System i Workspace Liberty interface by late 2026. Infor also emphasized cloud migration options and shared enhancements across its ERP suite, with version-specific updates like LX functional improvements, System21 3.3 features, and XA’s transition to Release 11.

Hiring

💼 Hire for curiosity, not your mirror image

If you’ve managed IBM i systems for over a decade at an organization ripe for modernization, here’s an uncomfortable truth: hiring someone just like you might be your biggest mistake.

It takes genuine self-reflection to recognize this. The tech landscape isn’t just changing—it’s compounding. What worked for the last ten years won’t work for the next two, let alone the next five.

What to prioritize instead:

Systems thinkers: People who understand how components interact within larger ecosystems
Natural curiosity: Those who ask the right questions rather than claim to have all the answers
Self-reflective learners: Individuals comfortable with continuous adaptation and growth
Leadership mindset: People who can navigate ambiguity and drive change
Cultural awareness: Those who recognize how technical changes impact organizational culture and approach transformation with sensitivity to existing workflows and team dynamics

Yes, IBM i background matters—it’s critical context. But don’t fall into the trap of hiring for exact technical skills match. That perfect skillset doesn’t exist, and even if it did, it would be obsolete within months.

The best hire isn’t someone who mirrors your experience. It’s someone who complements your gaps, challenges your assumptions, and brings fresh perspectives to problems you’ve been solving the same way for years.

The question isn’t “Can they do what I do?” It’s “Can they help us become what we need to be?”

Modernization

🔧 Legacy systems still dominate in 2025

Well, as this article highlights, we in the IBM i community are not alone, in that this survey reveals that:

62% of organizations rely on outdated software despite security risks.

The biggest blocker? Half say “the current system still works.” Yet legacy platforms trap data, block AI adoption, and drain budgets through rising maintenance costs.

Sound familiar?

If so, this article is a quick, and valuable read, that covers a number of topics, including why organizations delay legacy software modernization.

The time to act is now. Learn why the “wait and see” approach no longer works when it comes to IBM i modernization.

🔮 AI agents expose hidden gaps in legacy systems

This is one of the better articles that I have come across.

When integrating AI agents into enterprise systems of record, the real challenge isn’t technical connectivity—it’s bridging the gap between automated execution and human accountability. Agents can trigger actions that look valid but violate hidden business rules embedded in legacy platforms. The solution requires a policy-check layer that validates ownership, approvals, and context before committing changes, transforming AI from a fast executor into a trusted partner that respects organizational process and compliance requirements.

Open Source

Why Open-Source Software May or May Not be the Right Choice

Open-source software is often discussed in the IBM i community, but what does it really mean for your enterprise, and is it the right choice for your IBM i shop?

Definition:

“Open-source software is software whose source code is made publicly available so that anyone can inspect, modify, and redistribute it under a license that permits those freedoms.”

Security

🔓 AI companies leaking secrets on GitHub

While this does not involve the IBM i, it should make IBM i and RPG Developers take pause and take consider their security policies as they leverage GitHub.

Cloud security firm Wiz found 65% of 50 leading AI companies had leaked verified secrets like API keys and tokens on GitHub, often buried in deleted forks and developer repos. These leaks could expose organizational structures, training data, or private models.

Half of companies lacked proper disclosure channels. Wiz urges AI providers to run secret scans and implement disclosure processes, warning that speed cannot compromise security.

🔖IBM i Security Pages

If you have not done so already, or are new to the IBM i and it now falls under your responsibility, it is prudent that you bookmark this page to stay up to date on anything that relates to IBM i security.

As covered in a recent issue, a CTO for IBM i is sounding the alarm about security.

While the IBM i is noted for its secure nature, this security only holds true when proper precautions are implemented—make sure your organization has the right security measures in place to protect your systems.

Trends

🔮 The Best-of-Breed Future: Composable Systems Over Monolithic Platforms

Modern IBM i development is shifting toward best-of-breed composition—selecting specialized tools that excel in specific domains rather than relying on monolithic solutions.

This requires developers to embrace systems thinking—understanding how components interact within the larger organizational ecosystem. Successful teams are:

Leading with open-source development: Leveraging proven tools and frameworks from the broader development ecosystem
Prioritizing security at every layer: Building security into the development process rather than treating it as an afterthought
Integrating selectively: Choosing technologies that solve specific problems exceptionally well

The critical insight: start by taking a step back. Organizations must assess their needs, understand constraints, and design architectures that can evolve.

The IBM i platform’s strength has always been integration. In 2025, that means composing solutions from multiple sources—AI agents, modern development practices, or cloud services—to serve your organization’s specific needs.

Join

Contact us

If you are an RPG programmer looking to explore opportunities or a client who is looking for a talented IBM i professional, please contact us. We look forward to assisting you.

Do you know of someone who could benefit from Talsco Weekly? If so, please use the social media buttons to spread the word. Thank you!