Token Economy and AI Agents: The Final Hurdles in Enterprise Deployment

In this episode, Frank La Vigne sits down with his Red Hat colleague Christopher Newland for a deep dive into the evolving challenges and opportunities at the intersection of AI, open source, and enterprise technology. Fresh off attending both IBM Think and Red Hat Summit, Christopher Newland shares insights from two very different industry perspectives—executive strategy and hands-on engineering. Together, they explore the elusive “last mile” problem in AI adoption, the rise of agentic systems, the critical role of harnesses and runtimes, and why memory management is becoming the next frontier. Plus, they discuss the practical realities and future potential of tools like OpenShift AI, IBM Bob, and open source alternatives. Whether you’re a developer grappling with implementation details or a leader focused on ROI, this episode has something for everyone navigating today’s fast-changing AI landscape. Links Christopher on LinkedIn - https://www.linkedin.com/in/cjnuland/ Watch this episode on YouTube - https://www.youtube.com/watch?v=xlkPPt5YeY0 Time Stamps 00:00 Comparing IBM Exec and Red Hat Conferences 05:24 Challenges in AI implementation 06:56 Challenges in scaling microservices 11:38 Integrating AI with project management 14:23 Debate on AI model vs. harness 16:54 Discussing model evolution and limitations 22:54 Affordable Power BI Courses Bundle 25:19 Separating and managing runtimes 27:02 Using semantic routing for requests 30:15 Agent memory and compression basics 36:02 New AI approach and vision 38:49 Developing a multi-agent system 40:40 Importance of data chunking