Artificial Intelligence (AI) is one of the most frequently used terms in the data center industry. However, this brings the risk of losing sight of the bigger picture. That’s why Jasper de Vries, co-founder of Coolgradient, explains in an interview with the Dutch Data Center Association how AI can truly add value to data centers.
Everyone interacts with a data center every 18 seconds. It’s a quote from IDC’s Data Age 2025 that fascinates Jasper de Vries, co-founder of Coolgradient, because no one is aware of it. “A data center is very complex, but incredibly important for our society. At the same time, there is a lot of potential to optimize the data center with data and AI. That’s why René van Gompel (co-founder) and I started building the Coolgradient platform.”
The Coolgradient Platform
Today, the Coolgradient platform is used worldwide by various data centers. But how exactly does the platform work? “We use all data points from chillers, cooling towers, IACs, CRACs, CRAHs, UPSs, and PDUs used within the data center. We then use this data to better understand the data center. For example, we look at different conditions and determine the best settings based on them. We use machine learning to automatically generate this. The goal is to reduce the data center’s power and water usage and increase its reliability and resilience.”
One thing must be noted according to De Vries: Coolgradient is not a dashboard company. “Insights alone are useless if no concrete action follows. You need to know what to do to improve your efficiency. That’s why our platform’s AI focuses on 1) automatically detecting optimizations, and 2) automatically translating them into the right solutions for your specific data center. These are then presented as recommendations, and the platform provides quantifications. This last step is necessary so people immediately know what the solution is.”
Fruitpunch AI Challenge
To highlight the opportunities AI offers for gaining insights from data and acting upon it, Coolgradient has set up a collaboration with FruitPunch AI. This organization offers an online platform where AI for good challenges are provided. But how did this collaboration come about? De Vries explains, “We started talking with Sako Arts, CTO of FruitPunch AI, about collaborating to set up a challenge. The goal is to create broader awareness for sustainability within data centers. Additionally, we want to increase awareness within the market itself.”
According to De Vries, making a data center more sustainable can be done without expensive CAPEX. He paints a picture of data centers that traditionally replace a ten-year-old chiller with a newer version for sustainability. “Such a chiller costs about €400,000 each, and you already need around 10 on your roof. That costs an enormous amount of money, while you can already make significant steps in sustainability with the (sensor) data you have.”
The challenge AI For Greener Data Centers focuses on showing how AI can be effectively deployed within data centers, using various Machine Learning techniques such as Reinforcement Learning, to accelerate the implementation of an automated operating system.
This article was originally published by the Dutch Data Center Association and can be found here.