Glossary
On-premise AI
On-premise AI means operating AI software on infrastructure controlled by the company. Application hosting and model processing must be assessed separately: a locally installed application may still call external models.
Last updated: 2026-09-08
In a fully local setup, the application, search, data storage and models run on your infrastructure. Local application hosting alone does not guarantee that data stays local. Review model endpoints, telemetry, backups and optional external services.
Your organisation takes responsibility for access, updates, monitoring, backups and recovery. Assess available staff and infrastructure alongside the data requirements. Cloud and on-premise describe operating models; neither establishes security by itself.
Discuss the suitable deployment model and available introduction scope with chifty. On-premise is offered as an assisted pilot for selected projects. Local application hosting does not automatically imply local model processing; assess the enabled services separately.