Deployment and models
Your environment, or a dedicated instance we operate in the EU.
Either way, one customer per installation, and model usage runs on your own provider contract.
Two ways to run Polis
Choose where it runs. The controls are the same.
Models on your contract
The right model for each role, billed by your provider, to you.
Claude, via the Anthropic API or Vertex AI
Powered by Claude: Claude is the AI Developer’s coding engine.
Your choice of provider
- Claude
- Gemini (Vertex AI or the Gemini API)
- Amazon Bedrock
- Azure OpenAI
- OpenAI
-
Data classes per provider
Each model provider receives only the data classes you allow.
-
Billed by your provider
Model usage is billed by your provider, to you. Polis never resells model usage.
Polis stores no prompt or response content: usage records hold counts, codes and costs only.
In the product
Built for the people who run it.
-
Roles and two-factor authentication
Seven roles, two-factor authentication (TOTP) for every user, and step-up for sensitive actions.
-
English and German interface
The product interface is in English and German, with Swiss conventions.
-
Failover
If the coordinating service fails, another takes over within 20 seconds.
-
Envelope encryption
Secrets are protected by envelope encryption under a key-encryption key, with a check that a restore really works.
After the pilot
Operations features we plan next.
-
Operations features
Single sign-on (OIDC), monitoring and diagnostics, and licensing.
Book a demo
See it on your own workflow.
A 30-minute call: about 15 minutes of live product, including the pause-all switch, then your questions. Bring your CISO if you like.
What we show
- Roles that can’t overreach, with step-up for sensitive actions
- A model provider locked to a data class
- Approvals by a named person, in Slack or the panel
- The pause-all switch