Agentic AI Frameworks at IBM TechXchange Conference

What’s the next frontier for AI? On this episode of Six Five in the Booth, host Keith Townsend sits down with IBM's Alex Seymour, Product Manager for watsonx.ai, to explore Agentic AI—an innovative framework redefining how AI interacts with the world. Live from IBM TechXchange, they dive into how this groundbreaking tech is shaping industries and pushing the boundaries of AI innovation.Their discussion covers:

• The concept and benefits of Agentic AI frameworks.
• How IBM is implementing these frameworks in its products.
• The potential impacts of Agentic AI on various industries.
• Future developments and research in AI at IBM.
• Challenges and ethical considerations in advancing AI technology.

Learn more at IBM, watsonx developer hub, and Github developer repository.Watch the video below at Six Five Media, and be sure to subscribe to our YouTube channel, so you never miss an episode.

Keith Townsend:
All right. We're live, as you can tell, from the show floor of Xchange, IBM TechXchange. Sandbox 2.0, I love that name. I'm with Alex. Alex, IBM watsonx video. We're not going to get into the X. I want to talk to the developer audience. Talk to me about what's here for developers at TechXchange.

Alex Seymour:
Yeah. I think, just take a step back. Right? Developers are at the core of this next generation of applications. So within the .ai platform, we're focused on building and sharing a collection of services to support developers through that development lifecycle towards building this next generation of applications.

Keith Townsend:
All right. So I got to be blunt, we're, again, at IBM. I remember the days, the early days of watson, I had to call in to get someone to come in, build an application for me. I got a lot of value out of it, but I don't know if I would have called it a developer experience. How has that experience changed? What's the entry point for developers with IBM Cloud?

Alex Seymour:
Yeah, it's a really good question. So, we're focused on sharing collections of learning materials and assets to help developers start and accelerate the development of those applications. So we've got our Developer Hub, which contains all of the information and learning materials required to get started with the platform, as well as our Git repo, which contains a load of starting applications to provide the foundation for a lot of these applications.

Keith Townsend:
So we've talked abstracted. Talk to me about some level of detail. What type of applications are developers actually building with IBM?

Alex Seymour:
Yeah, it's a good question. So we see four core areas for applications in generative AI and enterprises. We've got applications focused on classification and summarization. So, asking models, "Right. We've got this transcript from our clients. Is this positive in sentiment, negative in sentiment?" And that kind of analysis. We're then seeing a lot of knowledge search in Q&A use cases. So, many clients today have massive collections of data and documentation about their business and want to expose that to their own internal stakeholders as well as their customers through a knowledge Q&A system. And then finally, we have the automation of business workflows. So we're talking a lot about agents and agentic behavior and how we can use those applications to automate the workflows that we do in our day-to-day role.

Keith Townsend:
So IBM with Granite are on a cutting-edge of small models that are labeled to me to customize these small models for my use cases. But talk to me about model selection in general. What's my options for models in the IBM development world?

Alex Seymour:
100%. So within our .ai platform, we offer a collection of third-party models provided by partners like Mistral and Meta with their Llama series of models. We've also released our own set of Granite models, open source, that are smaller focused models that are optimized for specific use cases. Now, the benefit of that with these smaller models is they're cheaper and they're more performant for these specific use cases.

Keith Townsend:
So bonus question, last question. Talk to me about the AI community within IBM. Where can I get help from my peer group?

Alex Seymour:
100%. So as I mentioned, we share a whole load of resources publicly and openly. If you want to connect with the AI community within IBM, we just released our developer-focused Discord channel. So that's where we'll be sharing updates and answering questions about how to build AI applications with the .ai platform.

Keith Townsend:
Alex, thanks a lot for joining me. Stay tuned for more coverage from the show floor of IBM TechXchange.