AI Platform
From Laptop-Based Analytics to a Scalable R&D Infrastructure | Hankook Tire & Technology
2026.03.23.
Hankook Tire & Technology had been strengthening its R&D capabilities through data analysis and modeling. At headquarters, around 30 data specialists were working across business areas on analysis, modeling, and algorithm development, using cloud-based tools including AWS and SageMaker.
But much of the actual work was still happening on individual laptops. A laptop-based setup is easy to start with, but as projects scale and data grows, both performance and collaboration hit clear limits. Internal high-performance computing units were mostly reserved for specialized researchers, making it hard for headquarters data specialists to access high-performance resources when needed.
To address this, Hankook Tire & Technology looked into a scalable analytics infrastructure. Rather than building large-scale infrastructure upfront, the company needed a setup that could start small and expand with project demand. To that end, it adopted Mondrian AI's Yennefer as the foundation for R&D data analysis and modeling.
■ Moving Beyond a Laptop-Based Environment
When analytics and modeling work runs on individual laptops, processing speed and runtime limits surface as projects grow. With larger datasets or more complex training pipelines, personal hardware can't support stable analysis or iterative experimentation.
Yennefer marked the turning point. Hankook Tire & Technology started with a small workstation setup and built a structure that could expand analytics environments per project as needed.
Data specialists were freed from the limits of personal devices and could run modeling and algorithm development in a more stable environment. Project-level experimentation and data processing became more flexible.
■ Project-Based Infrastructure on Yennefer
In R&D analytics, resource needs vary by project. Some focus on preprocessing and visualization; others demand more GPU resources for training and validation. A structure that adjusts to project demand works better than fixed infrastructure.
Yennefer let the team configure analytics environments per project and scale resources as needed. By leveraging cloud flexibility, it allocated resources efficiently and enabled more systematic resource management.
This approach reduces initial deployment overhead while staying ready for future demand. By starting at the right scale and expanding gradually, Hankook Tire & Technology improved the operational efficiency of its R&D environment.
■ GPU Sharing and Stronger Collaboration
GPUs play a critical role in data engineering and model training. When GPU access is limited to specific research groups, other data teams struggle to use high-performance resources.
Yennefer provided an environment for efficient GPU sharing and management. This improved resource utilization across teams and gave data specialists the compute they needed for modeling and algorithm development.
Yennefer's data storage and sharing features also strengthened collaboration. With a platform for sharing project data and results, analytics processes became more unified and knowledge sharing more active—boosting enterprise-wide data management and utilization.
After adopting Yennefer, Hankook Tire & Technology moved beyond a laptop-based environment and built a scalable R&D analytics infrastructure. Data specialists now work in a more stable environment, and GPU sharing and better collaboration have boosted analytics efficiency.
This case shows that Yennefer can be a scalable analytics platform even for organizations that can't deploy large-scale AI infrastructure at once. Through Yennefer, Mondrian AI helped Hankook Tire & Technology expand its analytics infrastructure step by step and strengthen its R&D collaboration and data utilization.
