Private / on-prem model
Running Coassemble's AI features against a private or on-premises model deployment, and how to arrange it.
Yes, by arrangement. Coassemble's AI features can be pointed at a private or on-premises model deployment instead of the shared service, so course content is processed inside your own environment. This is configured with our team as part of your setup rather than switched on from Settings, so talk to your account manager or contact our team to start it.
What a private deployment changes
- Where processing happens. Text generation, narration and image generation run against your own model deployment rather than the shared service.
- Which controls apply. Processing sits inside your own environment, retention rules and network boundaries, so it can be reviewed alongside the rest of your infrastructure.
- What your security team reviews. Because the model runs in your environment, AI processing can be assessed under the same controls as your other systems.
Everything else in Coassemble works the same way. Courses, sharing, reporting and the builder are unchanged.
Frequently asked questions
How do I set it up?
Contact our team. Setup involves confirming which model and deployment you're using, agreeing how Coassemble connects to it, and configuring your workspace. It isn't self-serve.
Which AI features work with a private model?
The same AI features are available. What changes is where the request is processed.
Is this the same as bringing my own key?
No. Bringing your own key runs Coassemble's AI features through your own account with an AI provider. A private or on-premises deployment runs the model inside your own environment. Both are arranged with our team, and which one suits you depends on whether your requirement is about billing and governance or about where processing physically happens.
Where is data stored by default?
By default, customer data is stored and processed on US-based infrastructure, currently in Oregon, USA. Region-locked servers in the EU, US or APAC are available on the Partner plan for teams with data residency requirements.