That’s when Identity-Aware Proxy with Open Policy Agent stops being theory and starts being oxygen.
Identity-Aware Proxy (IAP) lets you gate access to services based on who the user is, not just where they come from. Open Policy Agent (OPA) gives you fine-grained, programmable control over those decisions. Together, they form a zero-trust layer you can slip between your users and your infrastructure without rewriting your apps.
When you run complex systems, identity becomes your weakest link. An IAP ensures the request is coming from an authenticated user whose identity is verified in real time. OPA evaluates requests against centralized policies, written in the Rego language, to decide if the user should pass. This dynamic pairing means you can enforce team-specific access rules, compliance checks, and every “only if” condition your security model demands.
Setting up IAP with OPA starts with intercepting all requests at the proxy layer. The proxy authenticates with your identity provider—Google Workspace, Okta, Azure AD, anything that speaks OIDC or SAML. Once the user is known, the proxy calls OPA with a payload containing identity claims, request metadata, and any contextual signals you choose. OPA evaluates it all in milliseconds. Only if the policy allows does the request proceed to the service.