AI Provider

The AI Provider area lets your organization connect its own AI model to power the platform’s intelligent features — such as the quality and writing checks on requirements — using a provider and credentials you control.

For a regulated organization, the question “where does our data go when the AI runs” isn’t optional — it’s part of the diligence any auditor or security team will expect answered. Bring-your-own-provider is the answer: AI-assisted features run under your vendor agreement, your data-handling terms, and your governance, rather than through an opaque shared arrangement you can’t inspect. Organizations with contractual or data-residency constraints can point the platform at a private or regional deployment, keeping even the AI processing inside a boundary they control.

Bring your own model

Rather than depending on a single built-in model, you choose your provider — options include OpenAI, Anthropic, and Azure — and supply your own API key. You select the model to use and can set a custom endpoint URL where your deployment requires it, for instance when routing through a private or regional endpoint. A maximum tokens per request setting bounds the size of each call.

Why bring-your-own matters

For a regulated organization, control over where data goes and under whose contract matters. Using your own provider and key means AI calls run under your agreement and data-handling terms, your spending, and your governance — not an opaque shared arrangement. Organizations with a private model deployment can point the platform at it directly through the custom endpoint.

Where the AI is used

The configured provider backs the platform’s AI-assisted capabilities, including the requirement quality analysis described in Quality scoring. Configuring it here, at the organization level, ensures every project’s AI features run on the model and terms your organization has chosen.