Based on the skyrocketing demand for artificial intelligence, having a holistic, iterative and automated tool for AI model testing, such as Podman AI Lab, is vital. 

As an open-source extension for Podman Desktop, Podman AI Lab is a game-changer since it enables developers to build, test and run AI-powered applications in a local environment using containers, according to Ignacio Riesgo (pictured, right), senior director of developer marketing and strategy at Red Hat Inc.

Red Hat’s Kevin Dubois and Ignacio Riesgo talk to theCUBE about how Podman AI Lab streamlines model testing.

“We have evolved AI, security, connectivity — everything now is connected,” Riesgo stated. “We are starting to move the conversation within developers because now Podman Desktop has Podman Desktop AI Lab. That is how to test your models, the LLMs inside your own machine. At the same time, we have a Developer Hub that is also evolving. That is how you enable new developers to come.”

Riesgo and Kevin Dubois (left), senior principal developer advocate at Red Hat, spoke with theCUBE Research’s Savannah Peterson and Rob Strechay at KubeCon + CloudNativeCon NA, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how Podman AI Lab empowers developers to leverage AI capabilities easily and efficiently. (* Disclosure below.)

How Podman AI Lab aligns with developers’ needs

Given that Podman AI Lab democratizes access to AI, it makes it easier for developers to experiment and learn while maintaining control over the development process. This is because the tool enables developers of all levels to explore large language models and generative AI features without requiring extensive AI expertise, according to Dubois. 

“I think the main goal also is to make sure developers are happy with everything,” he said. “We have, for example, Podman AI Lab, where you can just run a model on your local machine. Then InstructLab, which is really interesting to train models without being deep experts into these technologies. It spans the whole spectrum. Even for Java developers, we’re investing in Quarkus as a Java stack where you have really smooth integration with LangChain4j.”

By incorporating secure code analysis, organizations not only improve the security posture of their software, but also foster a culture of responsibility and quality in development practices. As a result, Red Hat Profile Analyzer helps with this objective since it enables developers to understand an enterprise’s risk profile, according to Riesgo. 

“There is another tool, not only the Signer but also the Analyzer,” he said. “You have a tool that analyzes how risky the code is, how risky are the dependencies that you are taking in order to build your app. Having the capacity to analyze your pipeline is also critical. Not only having the capacity to sign that you are doing the right things, but where are those sources coming from?” 

Effective interconnectivity management is the backbone of a robust multicloud strategy since businesses can harness the full potential of diverse cloud platforms while minimizing risks and inefficiencies. As a result, Red Hat has set the ball rolling through its Interconnect platform, Dubois pointed out. 

“We started with Kubernetes and cloud environments — we’re going all into one environment,” he said. “What we’re working on is a higher level interconnectivity component where it’s more for the developers to say, ‘Hey, I want to run this here.’ We connect it together, it’s like API management, but beyond that. It’s interconnectivity management because we can also do things that we could do in API management in terms of rate limiting, making sure somebody has authentication and authorization by different personas or by different components.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of KubeCon + CloudNativeCon NA: 

(* Disclosure: Red Hat Inc. sponsored this segment of theCUBE. Neither Red Hat nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE