Viking Analytics was founded in 2017 in Gothenburg, Sweden by a group of engineers and researchers who saw the future of maintenance differently.
With backgrounds in machine learning, signal processing, and enterprise software, we knew what AI could do, but also what it couldn’t. Instead of building another black box, we focused on creating a system that works with analysts, helping them find what matters faster and deliver better answers, not just more alerts.
At the time, most predictive maintenance tools relied on threshold alarms or rigid fault libraries. Meanwhile, AI was overpromising. With rule-based diagnosis tools barely hitting 30-40% accuracy and prescriptive systems performing even worse. The reason? Vibration data alone isn't enough. Machine behavior depends on context: operating conditions, machine train dynamics, process interactions, even plant layout. And much of this insight lives inside the minds of human experts.
So we built MultiViz: an AI platform that doesn't try to replace the analyst, but supports them. We designed it to make their work faster, sharper, and more scalable. Our goal was to help them focus. Not on sifting through hundreds of waveforms. Instead, on solving real problems. MultiViz points to what’s unusual and why. The analyst decides what matters.
Over the years, we’ve partnered with manufacturers, service providers, and OEMs across Europe, North America, and Latin America. We’ve helped teams monitor thousands of assets without drowning in data or losing control. Along the way, we also created Daeploy, an open-source tool for deploying Python models, because we believe powerful tools should be easy to use.
We’re proud to be recognized by AI Sweden as one of Europe’s leading AI innovators. More than that, we’re proud of the teams we work with; they're the reason we exist.
And we’ll keep building for them – until no machine failure comes as a surprise.
