Security & deployment
Athena Platform runs only in your environment. Data, model calls and audit records stay within your control.
Two ways to deploy
On-premises
Installed in your data center or private infrastructure.
Customer-provided cloud
Deployed into your own cloud account or environment, under your control.
We do not offer SaaS and we do not host your data.
Built-in controls
- Server-side identity and ownership
- Users, projects and resource ownership are derived from authenticated identity and server-side relations, never from IDs supplied by the client or a model.
- One gate for model calls
- All generation and embeddings go through the LLM Gateway, which applies personal-data and cloud-use policies.
- Risk-tiered confirmation
- Irreversible actions require strong confirmation or step-up and are audited.
- Internal services stay internal
- The assistant runtime, Gateway and model endpoints are not exposed directly to the internet.
- Separate speech nodes
- Speech recognition and synthesis run on dedicated nodes, managed apart from other services.
Delivery & operations
Delivered as pinned image versions with a release manifest and migration ledger, together with rollback and disaster-recovery runbooks.
What we do not claim
This site gives no availability, recovery-time or throughput figures and claims no regulatory certification. Those must be verified for your environment.