If your mission is to help organizations add AI agents to accelerate processes, you need to start with the basics. This means making data available for AI use. As Niels Zeilemaker, global CTO at Xebia, explains, Agentic AI scales with the strength of your data.
“If you don’t think about that, you can build the best agent, but it won’t find the right data. You’ll probably misinterpret the data and combine different fields in the data that shouldn’t be connected,” Zeilemaker explains. “And these mistakes aren’t necessarily the agents’ fault; they’re the Foundation’s fault for not being ready for an AI agent.”
One area of particular consideration is data cataloging, Zeilemaker points out. This is not a new concept, but it changes things for agents. “When you’re setting up a data catalog for an organization that’s comprised entirely of humans, there’s always a fallback,” he says. “If something isn’t well-documented, you can pick up the phone and walk up to your colleague and have a kind of back door to say, ‘What should we do with this particular data set?’
“The agent has no such backdoor. It has to rely on the data catalog and what it says. If that description is wrong, the agent won’t work.”
Xebia is focused on helping organizations transform their AI strategies into production-ready solutions and drive real transformation faster. The company’s core values include putting people first and uncompromising quality, but perhaps most importantly, as Zeilemaker sees it, is sharing knowledge, including at events like TechEx Global North America, where Xebia has participated.
“I think sharing knowledge is very important to us, and sharing knowledge allows us to stay a little ahead of the curve and quickly adapt to new changes in the market, because everyone has an eagerness to learn new things and share what works and what doesn’t,” says Zeilemaker. “By putting a lot of effort into sharing this knowledge and innovation, we also try to pick some areas where we want to be an authority.”
Data and AI is clearly one of those areas. At the AI & Big Data Expo, Zeilemaker explained to attendees how to build on this AI foundation and unify fragmented data environments. This was an honest account of how a combination of dedicated AI agents and expert engineering compresses 12-24 month timelines into fixed-price, milestone-only engagements.
The overarching thread for this is what Xebia calls Agentic Data Foundation (ADF), which extends the data platform to hosted agents and leverages them for both customer-facing use cases and internal processes. While there is always great interest in migrating from legacy platforms to modern platforms, Xebia is seeing more and more customers looking for a faster and more reliable approach to migrating to their data platform. Zeilemaker says this is where consultants and customers co-develop solutions.
“The agent has to trust the data catalog and what it says. And if that description is wrong, the agent won’t work.”
“After doing the migration the old-fashioned way and speeding up some parts with LLM coding, we are now integrating this into our data platform and taking advantage of the additional context it can provide to further accelerate the migration,” he says.
That accumulated experience is what shaped Xebia Axis. Agentic Data Foundation is Xebia’s answer to helping enterprises make their data AI-enabled.
Another weapon in Xebia’s arsenal is Xebia ACE: AI-Native Software Engineering, a framework that embeds AI throughout an organization’s software development lifecycle (SDLC). Done right, delivery can be up to 40% faster and traditional transformation costs can be reduced by up to 70%.
Zeilemaker says Xebia ACE is especially useful for large companies that “want to stick to a certain governance or way of working while still executing their SDLC.” There’s a bigger picture here. Zeilemaker uses vibecoding as an example. “If you think about vibe coding, anyone can create apps, but no one is actually trying to push these apps into production,” he says. “With ACE, you still get many of the benefits of LLM acceleration, but with the same quality end result as before.
“If you are looking to switch to using LLM in your coding, Xebia ACE provides a very nice framework to use without the risks and drawbacks of running a dark factory LLM and hoping for the best, and losing a bit of control and governance in the process,” adds Zeilemaker.
For companies, that control is important. Because so much code is generated, AI-driven SDLCs can become a security weakness due to vulnerabilities. Zeilemaker maintains that this is something the industry still needs to understand to some extent, but he notes with interest Anthropic’s recent move to release a pull request reviewer.
“It’s interesting and we’ll probably see more of it,” he says. “We have a very long pull request review, which applies every time we try to run a new product release, and we add very senior team members to the process in the form of LLMs to do a kind of third-party review.
“I think this is an interesting perspective on what we’re going to see more of in the future.”
Ultimately, no matter what stage your organization is at, from assessing data readiness to preparing to build, Xebia can help you get the foundation right and create transformation on top of it.
Photo by Fabio on Unsplash
Want to learn more about AI and big data from industry leaders? Check out events in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other major technology events. Click here for more information.
AI News is brought to you by TechForge Media. Learn about other upcoming enterprise technology events and webinars.
