Founded in 2001 and headquartered in the Netherlands, ITQ is a leading, multi-cloud company committed to delivering innovative and high-quality IT solutions. Over the years, ITQ has expanded across Europe with a workforce of over 250 employees. The company provides support, consultancy and managed IT services, specializing in digital workspaces, hybrid cloud, cloud native platforms, cyber-resilience and AI.
At-a-Glance
ITQ is redefining human-machine interaction by training AI models to recognize human gestures and translate them into real-time robotic commands. By leveraging SUSE AI and SUSE Rancher Prime, ITQ has created a responsive, gesture-controlled robot dog. With a secure and flexible foundation, ITQ has cut time-to-value and unlocked an agile methodology that can be used on future projects.
Enabling secure AI for the enterprise
Always aiming to keep pace with market innovation, ITQ has expanded its solutions portfolio in recent years. Recognizing the rise of machine learning (ML) and artificial intelligence (AI), the company has invested heavily in AI expertise, including their own AI research and development (R&D) lab.
While interest in deploying AI solutions to boost productivity and streamline workflows is increasing, many companies do not have the required infrastructure and skills to run and maintain large language models (LLMs) in-house. Simultaneously, there are growing concerns regarding the security risks of uploading training data — potentially containing proprietary knowledge — to the public cloud. For those in highly regulated fields like government services, digital sovereignty principles mean that leading hyperscalers based in other nations are not always an option. Furthermore, while cloud infrastructure can offer significant cost savings over on-premises infrastructure, many businesses encounter significant cost-control challenges when using the public cloud for AI services.
As a result, enterprises are looking for more flexible ways to explore and deploy AI. To support its customers in enabling AI-powered workflows rapidly, securely and cost-effectively, ITQ needed a new solution that could be seamlessly deployed on-premises or at the edge of the network.
“With SUSE AI we can help customers implement initial use cases and drive ROI in a couple of months or a few weeks even, using the same agile methodology we used to develop Q9. Building a similar open source solution from scratch can take up to 18 months, and in that time the hardware sits unused, which is just a loss of investment.”
Johan van Amersfoort
Chief Evangelist, AI Lead
ITQ
Why SUSE solutions?
To prove that AI could be deployed securely and rapidly at the edge, ITQ developed a high-profile proof of concept: a gesture-controlled quadrupedal robot dog named Q9. The inspiration came from an internal innovation project led by Johan van Amersfoort, Chief Evangelist and AI Lead at ITQ.
Having recently developed a home automation system using computer vision and AI to manage his garden sprinklers, van Amersfoort recognized that the same underlying logic — flexible AI training and lightweight edge deployment — was the answer to the market’s cost concerns. He realized he could use a quadrupedal robotic dog as a mobile edge device to showcase these capabilities in real time. As an initial use case, van Amersfoort aimed for a model that could interpret human hand gestures as commands to perform actions.
Leveraging its deep expertise as a SUSE Diamond Partner, ITQ worked with SUSE AI experts to design a custom stack for AI-augmented robotics. The team chose SUSE Rancher Prime and SUSE AI as the foundation, valuing their ease of use and SUSE’s commitment to flexible, open source architecture.
Working with co-developer Sander Harrewijnen, Technologist and Developer Advocate at ITQ, van Amersfoort trained and deployed the first prototype in just four weeks. Utilizing a CI/CD methodology with Argo CD integration, the team trained the model on a dataset of images of 15,000 hands. Van Amersfoort continued to work closely with SUSE over the course of two months to refine the concept.
To ensure the near-zero latency required for real-time interaction, the AI model identifies 21 “landmarks” on a human hand to verify gestures. The robot dog reacts to the identified gesture by performing the corresponding action. The AI model runs on an edge device powered by K3s, the lightweight Kubernetes distribution, managed via SUSE Rancher Prime.
Building on this success, ITQ is now developing a larger LLM that supports both visual and textual input, using the SUSE AI virtual assistant stack to give Q9 a voice. Hosted on Dell servers with NVIDIA GPUs at ITQ’s AI lab, the LLM will allow Q9 to capture images at events, interpret them through a custom-built API and autonomously create topical posts for social media.
The impact of SUSE solutions
Accelerates AI adoption by 12x
Q9 demonstrates how rapidly organizations can adopt AI across a variety of use cases with SUSE AI. By utilizing streamlined and automated workflows, the SUSE tech stack curated by ITQ is designed to accelerate time-to-value for any edge-based AI project.
Previously, developing a self-service GPU-as-a-service AI solution for a customer took ITQ around three months. With SUSE AI, the company can deliver similar results in just a week, boosting delivery speed by 12 times.
“With SUSE AI we can help customers implement initial use cases and drive ROI in a couple of months or a few weeks even, using the same agile methodology we used to develop Q9,” says van Amersfoort. “Building a similar open source solution from scratch can take up to 18 months, and in that time the hardware sits unused, which is just a loss of investment.”
Enables secure, flexible AI deployment
By leveraging SUSE AI, ITQ and its customers can deploy AI across any environment — from centralized data centers to a robot dog at the edge — maximizing both agility and cost-effectiveness. With a range of LLMs, tools and components to choose from, customers can tailor solutions to their unique needs, from customer support chatbots to data analysis platforms and more.
“SUSE AI is a complete, CNCF-supported solution that works from the get-go, whether on an enterprise-grade infrastructure or on an ARM device at the edge,” confirms van Amersfoort.
This flexibility is central to security and digital sovereignty. By avoiding vendor lock-in and reducing reliance on the public cloud, customers can maintain full control over their data. Furthermore, because the solution is built on a verified, auditable supply chain, it meets the regulatory requirements for even the most sensitive industries.
“With SUSE, you have the advantage of open source while having an enterprise behind you to support security,” adds van Amersfoort. “SUSE AI is secure and sanitized by default, and you never have to worry about certificates or security patching.”
Centralizes Kubernetes management across environments
Using SUSE Rancher Prime, ITQ centrally manages all of its Kubernetes clusters, from the edge to the cloud and from traditional Kubernetes to K3s and RKE2. The user-friendly interface simplifies cluster management, facilitating productivity and agility. What’s more, SUSE Rancher Prime enables ITQ to quickly spin up new clusters, supporting scalability in rapidly changing markets.
Supports real-world AI use cases
Today, as ITQ’s “Chief Technology Wooficer”, Q9 can recognize numerous human gestures, performing tasks from a simple wave to a backflip. However, these interactions are just the beginning. ITQ envisions a future where Q9 serves as a sophisticated robotic guide dog — complete with its own personality — supporting individuals who may be unable to own a living service animal due to allergies or other constraints.
With responsive SUSE experts on hand to help ideate and refine solutions and resolve any issues, ITQ is confident in its ability to lead more customers through their AI journeys.
What’s next for ITQ?
In the near future, ITQ plans to deploy SUSE Observability on Q9 to gain deeper visibility into the robot’s telemetry and optimize its performance. To further streamline the development lifecycle, ITQ is also evaluating the integration of Kubeflow, replacing a proprietary ITQ training pipeline with a standardized, open source framework, ensuring the project remains scalable, transparent and fully aligned with current AI best practices.
“SUSE’s interest in this project and their collaboration really stands out to me,” concludes van Amersfoort. “Q9 is a constant R&D project, like a Swiss Army knife for showcasing the technology. As new features are added by SUSE, we will be keen to explore them and run the AI platform to its full potential.”