Passwordless Authentication for RADIUS, VPN, and Windows Desktop Access
People may unlock a workstation, connect through a VPN and open a federated application within minutes. Those are different access paths. A useful procurement matrix identifies the point where authentication happens, the component that enforces access and the combination that needs to be tested.
Start with the access surface
| Access path | NoPass scope for evaluation | Evidence to collect |
|---|---|---|
| Windows Desktop Unlock | A configured workstation unlock path for enrolled users | Windows edition, managed device configuration, enrollment and recovery |
| Initial Windows sign-in | Separate access path | First boot and signed-out behavior in the target endpoint configuration |
| RDP | Separate remote session path | Gateway, client and remote host behavior for the stated topology |
| RADIUS / VPN | Configured RADIUS access and VPN paths | Gateway make, firmware, RADIUS policy, timeout and failure handling |
| OIDC and SAML | Configured federation paths | Issuer, client or service-provider registration, claims, logout and access policy |
| Keycloak | Configured identity-broker or identity-provider integration | Adapter, protocol, version, registration and error handling |
| Entra / Microsoft 365 | Configured Microsoft identity workflows | Tenant topology, where the factor is enforced, session and recovery behavior |
Configured Windows Desktop Unlock, RADIUS/VPN, OIDC/SAML, Keycloak and Microsoft identity paths have operated in an internal production environment. Use that experience to choose evaluation routes. Confirm the specific VPN gateway, Windows access point or Microsoft tenant policy separately. If your requirement names Entra External MFA registration or an exact Conditional Access authentication strength, ask for configuration evidence for that requirement.
A buyer checklist for one pilot
- Pick one access path and define what success looks like. For example, unlock an already enrolled, locked managed workstation under the intended network conditions.
- Identify the enforcing system. The gateway, endpoint or application decides whether to grant access. Record the outcome when the authentication service is unreachable or the user declines.
- Record the version tuple. Include service, client, endpoint, gateway or identity provider and the policy changes needed for the pilot.
- Test recovery with representative operators. Cover a lost device, a temporary connectivity failure and account removal. Record how the service desk can help and how the enforcing system handles the result.
- Extend the matrix only after a result exists for the next access path. A working workstation unlock or a successful federation sign-in should be cited with its own evidence and configuration.
How to interpret a result
A passed scenario should name the environment, the outcome and the component that made the final decision. If the gateway times out, use the record to distinguish the delivery problem from the gateway policy result. For federation, make sure the application receives the expected user mapping, handles logout and observes the policy the team configured. Verify that the help desk can recognize each failure and has a documented route for restoring access.
Bring one versioned access route to an integration discussion. Related pages: NoPass Workforce, Desktop Unlock and Microsoft Entra authentication methods.
