Configure AI Providers

This guide explains how Tenant Administrators manage AI provider settings, Bring-Your-Own-API (BYO_API) keys, and specialized task models in Administration > AI Settings (/settings).

---

Operating Modes

1. Default Hosted Platform AI: Standard AI capabilities powered by platform-managed inference.
2. Bring-Your-Own-API Key (BYO_API): Utilize your organization's direct API agreements with supported providers (OpenAI, Anthropic, Google Gemini, Azure OpenAI, Ollama, Replicate).

---

Configuring BYO API Keys

1. Navigate to Administration > AI Settings (/settings).
2. Set AI Mode to BYO_API.
3. Select your Provider (e.g., OPENAI, ANTHROPIC, GEMINI, AZURE).
4. Enter your API Key and optional organization/endpoint parameters.
5. Click Verify & Save. The platform performs a lightweight validation check and securely encrypts your credentials in the platform security vault.

---

Assigning Task-Specific Models

You can configure specialized models for distinct AI tasks:

• Extraction Model: Optimized for document structure parsing and evidence fact extraction.
• Assessment Model: Configured for design adequacy and control evaluation reasoning.
• Assistant Model: Default conversational model powering the slide-over OrviQ Assistant.