Briefer’s enterprise plan allows you to configure your self-hosted deployment to use AI models in your own AWS account. This way, you can keep your data private while still benefiting from Briefer’s AI capabilities.

Briefer’s private AI setup uses Amazon Bedrock, a service that allows you to run AI models in your own AWS account.

Briefer’s private AI setup is only available on the enterprise plan.

​
Setting up private AI

To set up private AI in your self-hosted Briefer deployment, follow these steps:

1

Ensure AI is enabled for your deployment

Ensure that your Briefer deployment has AI enabled. You can do this by toggling the ai.enabled setting in your Briefer configuration file.

2

Enable the desired models on AWS

Enable the AI models you want to use in your AWS account. You can do this by following the instructions in the Amazon Bedrock documentation.

We recommend using Anthropic’s models, preferably Claude v3.7 Sonnet or later.

3

Configure your service account's IAM role

Update the IAM role used by the Briefer API’s service account to allow it to access the AI models you enabled in the previous step.

Briefer needs the following permissions to use Bedrock models:

"bedrock:InvokeModel"
"bedrock:InvokeModelWithResponseStream"
"bedrock:ListFoundationModels"
4

Select the desired model (optional)

In Briefer, go to the “Settings” page and select the desired AI models in the “AI models” section.

By default, Briefer will use the latest version of Anthropic’s Claude model.

Granular permissions