Where inference runs

Spaces is built for organisations that need to know where their data goes — and where it does not.

Compute in our own facility

Answers are generated on Empathy.AI's GPU infrastructure — dedicated GPUs in the Empathy.AI headquarters, a net-zero bioclimatic building in Asturias, Spain, powered by an on-site solar canopy. Processing runs on our own hardware, not on public cloud or shared third-party infrastructure.

Private infrastructure

Answers are not generated through third-party cloud APIs or shared model endpoints that could expose your content. They are generated on open-source models we run and maintain ourselves.

No data egress

Your Knowledge, conversations, and Controls stay inside your Space perimeter. Spaces does not send your material to external services for processing or training.

No shared models trained on your data

Conversations are processed privately and never used to train any model. This is a core product commitment, not an optional setting.

How this Space answers is governed in Controls. This page covers where processing runs.