Add Bring Your Own Model Capability

[tejugang]

Enable users to integrate their own hosted inference models by providing an endpoint, authentication token, and model name. This allows flexibility for teams to choose and manage their own models instead of relying on a fixed one.

Acceptance Criteria:

• Configuration support for custom model endpoint, token, and model name
• Documentation on how to connect to external inference services
• Validation to ensure the provided model endpoint is functional