What is an in-tenant AI deployment?
An in-tenant AI deployment runs the AI inside the customer's own cloud tenant, such as their Azure environment, rather than in the vendor's cloud. Deal data never leaves the firm's environment, access flows through the firm's existing permissions, and the firm's own security controls apply. For financial firms handling confidential deal information, security teams increasingly treat this as a requirement, not a preference.
The default SaaS model asks a firm to ship its most sensitive material, live deal email, buyer conversations, mandate details, to a vendor's servers and trust the vendor's controls. For a deal firm, that is often where the security review ends and the deal dies.
In-tenant means the software comes to the data instead. The system is deployed inside the firm's own cloud subscription, reads email and documents through the firm's existing permission model, and processes everything within the boundary the security team already governs. There is no second copy of the firm's deal history sitting in someone else's cloud.
The questions that separate real in-tenant deployments from marketing: where does processing happen, what data leaves the tenant and when, whose identity and permission system gates access, and can the firm's own team audit and shut it off? A deployment that answers all four inside the firm's boundary is in-tenant.