Private GPU cloud for financial servicesConfidential data, dedicated hardware, a European operator.

What matters in financial services

Models trained on transaction, customer or risk data carry that data's sensitivity with them. Vendor due diligence needs clear answers on location, tenancy, operator and access; finance teams need a cost they can budget.

• EU data residency. Customer data is processed and stored on Sapience systems in Bratislava, Slovakia. Sapience does not replicate customer data outside the EU; backup and replication arrangements are defined in the contract.
• Single-tenant isolation. Dedicated compute, storage and network with hardware-level segmentation; private clusters on request.
• Encryption. TLS 1.3 for all data in transit, AES-256 for all data at rest, encrypted NVMe on every node.
• European jurisdiction. Operated by European entities under European law, with no dependency on US hyperscalers for compute, storage or networking.

What Sapience provides

• Single-tenant bare metal with direct hardware access: your stack, your access controls.
• Written answers to vendor assessments from the people who run the infrastructure.
• Reserved capacity for predictable cost, priced per GPU-hour for a term: see pricing.

Model development usually starts with fine-tuning and inference on RTX PRO 6000; larger training runs use HGX B300.

Compliance status, stated plainly

A certification is listed as in progress until the audit is complete. Ask for the security overview and data-processing terms through the security page.