SUSE wants a piece of the AI cake, too

SUSE, the venerable Luxemburg-based open-source company, has long been a household name in IT circles in Europe, but it’s never quite managed to capture the U.S. market, where competitors like Red Hat and Canonical are far better known. Yet, just like as in the cloud world, where a lot of players are hoping for AI to reshuffle the playing field, so, too, does SUSE hope that AI will give it a new entryway into the U.S. market — together with its recent moves to more directly challenge its competition. The company on Tuesday is announcing its AI strategy and SUSE AI solutions, a new vendor- and LLM-agnostic generative AI platform.

Ahead of the announcement, I exclusively spoke with SUSE CEO (and ex-Red Hat executive) Dirk-Peter van Leeuwen and Pilar Santamaria, the company’s recently appointed VP of AI, about the new service and SUSE’s overall strategy with regards to AI — but also open source in general.

“The vision of SUSE is to bring the infinite potential open source to the enterprise,” van Leeuwen, who became SUSE’s CEO in March 2023, told me. “We do believe that this open-source model is giving us an infinite potential; it just keeps evolving faster than any other development model because it’s exponential. It’s extremely iterative. And because it’s open, people use this for many different things than what the original developer wrote it for. We’ve seen this all in the internet, with AI, with all the things that are happening around us. It’s all driven through open source. But of course, as we all know, for enterprise customers, you need a little bit more than just access to code. You need support, you need security, safety. And most importantly, you need to be sure that your product will be supported for long term.”

The question of long-term support is what got SUSE to fork CentOS and support existing customers when Red Hat changed its development model for the popular Linux-based operating system last year. That, van Leeuwen said, has resulted in a “tremendous uptick” from former CentOS users who are migrating to SUSE’s fork. “Customers truly like to tap into this opportunity of switching vendors without switching software,” he said, comparing it to mobile phone users simply switching out their SIM to get on another network. “In software, you could never do this except with open source, and that’s really what I wanted to achieve with this offering.”

He also noted that a lot of these enterprises then take a look at SUSE’s overall portfolio, which besides its core Linux offerings also includes Kubernetes service Rancher and security service Neuvector, which the company acquired under former CEO Melissa Di Donato. During a time where enterprises are looking to consolidate platforms, that’s a major advantage. But SUSE itself also went through its own slew of ownership changes over the years and that hasn’t necessarily helped it position itself in the market.

“SUSE has been, and still is and always has been, an amazing company,” he said. “But the downside for SUSE as a company has been that it’s gone through quite a few acquisitions. And when you go through these acquisitions, you get a new management, a lot of stuff gets reset, and the world is moving very fast, right?” SUSE, he said, has always done well in its work with SAP, which helped it grow in the European market, but the U.S. has remained a challenge.

“In the U.S., SUSE has never really reached brand recognition. That is something we’re working on as well. Because U.S. customers in particular, are in many cases not even aware of the existence of SUSE. We’re hard to pronounce for U.S. customers. And so there are things for us to work on. But they’re not the hardest thing to do, because we have the products and we have the solutions and customers like this,” he said.

He stressed that Rancher is already a strong brand in the U.S., so the company plans to connect that closer to the overall SUSE brand and get those customers to look at more than just its Kubernetes offering.

AI is obviously the other area where SUSE thinks it has an opportunity to grow. At its core, the company sees itself as an open-source infrastructure player — and the next frontier there is supporting AI workloads, after all.

The new SUSE AI solution — which itself is open source, of course — is squarely aimed at helping its customers put AI workloads into production, and to do so in a secure and privacy-first way. It’s worth noting that it’s not a training solution but is meant to help businesses use their own models or open-weight foundation models like Meta’s Llama.

“Many companies cannot really use generative AI because they find that they have to give their data to third parties. Basically, they don’t feel they can drive in AI — and if you don’t drive, you’re the data. That’s it,” SUSE AI VP Santamaria said. And even if they don’t mind that, then a lot of companies face compliance issues because a vendor may not be able to guarantee where in the world the data is processed.

Santamaria argues that until now, there wasn’t an open-source solution on the market that gave enterprises the freedom to run these LLMs in their own cloud or virtual private cloud — combined with the access controls and security solutions they need. “This is the first solution in the market with these components, totally turnkey, and deployed in minutes, not in days,” she said.

She stressed that the company thinks that users must have the freedom to deploy the models of their choice, maybe fine-tuned or augmented with a company’s own data using retrieval-augmented generation. But at the same time, the industry is moving so quickly that many users also don’t want to lock themselves into a single vendor that may or may not be at the forefront of what’s next.

The idea here is for the solution to be modular, allowing people to select the vector database of their choice, for example, to build a solution that best suits their needs.

One of those customers is Fujitsu. “Generative AI is helping to unlock innovation within our world. Our customers’ employees already use generative AI in their private lives and naturally want to use this technology at work too. With our solution, they can do this in a secure and protected setting,” said Udo Würtz, Chief Data Officer, European Platform Business at Fujitsu. “As a trusted partner, SUSE supports us with our genAI product strategy through their collaboration, expertise and commitment to choice for customers.”

SUSE’s AI solution is now available as an early-access program