Frequently Asked Questions

Everything you need to know about Sapience AI's EU-sovereign GPU infrastructure

What is Sapience AI?

Sapience AI is an EU-first GPU-as-a-Service (GPUaaS) platform. We provide low-latency, EU-sovereign GPU infrastructure for ML training, inference, and high-performance computing workloads. From single-GPU experiments to multi-node superclusters — everything is built and operated exclusively in Europe.

Is my data GDPR compliant and EU-sovereign?

Yes. Data sovereignty is our foundation. Every workload runs in a fully isolated, single-tenant environment within the EU. Compute, storage, and networking are segmented at the hardware level per customer. We operate under full GDPR compliance and are pursuing ISO 27001 and SOC 2 Type II certification. For regulated industries like healthcare, fintech, and public sector, our EU-only architecture eliminates cross-border data transfer risks.

Do you support multi-GPU and multi-node training?

Yes. Multi-GPU nodes are available with up to 8 GPUs per instance for data-parallel or model-parallel training. For distributed workloads across multiple nodes, we provide 10 Gbps inter-node networking with 100 Gbps InfiniBand options for large-scale clusters. Contact our team for custom cluster configurations.

How do I get started?

Request early access from our homepage or email sales@sapienceai.eu. Once approved, you get console access to spin up GPU instances within minutes. For enterprise needs like dedicated clusters, custom SLAs, or volume pricing, use our contact form and select "Long Term Rental" or "Custom Build." We typically respond within one business day.

Where are your data centers located?

All Sapience AI infrastructure runs within the European Union. Our GPU clusters are deployed in SITEL and VNET Tier 3 data centers in Bratislava, Slovakia, with expansion planned to Frankfurt and Amsterdam. Your data never touches a server outside the EU — that's guaranteed by architecture, not just policy.

Can I use my own Docker images and ML frameworks?

Every instance provides full root SSH access. All instances ship pre-configured with CUDA 12, cuDNN, NCCL, PyTorch, TensorFlow, and JAX — ready to train out of the box. You can also deploy any custom Docker container, Conda environment, or additional framework via pip or apt. JupyterLab is accessible directly from the console for interactive development.

What security measures do you have in place?

Every customer runs in a single-tenant environment with hardware-level isolation of compute, storage, and network. Our data centers operate at Tier 3 standard with redundant power, N+1 cooling, biometric access, and 24/7 on-site security. All data is encrypted in transit (TLS 1.3) and at rest (AES-256). We are actively pursuing SOC 2 Type II and ISO 27001.