SAML assertions are being rejected
Usually a rotated signing certificate, a mismatched entity ID, or XML canonicalisation that the two sides disagree on. The browser trace tells you which.
Assertions verified and accepted
Working with SSO?
Single sign-on fails in small, specific ways: one attribute named wrong, a certificate that rotated last week, a redirect URI that does not match by one character. We read the actual assertion and the actual error, then make them agree.
What it is
Single sign-on lets users log in once and reach multiple applications without entering a password at each one. The two dominant protocols are SAML 2.0, an XML-based standard that has run enterprise B2B federation since 2005, and OpenID Connect, a JSON layer on top of OAuth 2.0 that now carries most consumer and modern business logins.
SCIM 2.0 handles the user and group provisioning side, so accounts appear and disappear in the application as the identity provider changes them. The identity providers people actually run include Okta, Microsoft Entra ID, Google Workspace, Auth0, OneLogin, JumpCloud, Ping Identity and the self-hosted Keycloak. Hawkins.Tech works on both sides of an SSO integration: the application that accepts the login and the enterprise customer configuring their directory.
The honest version
No technology is all upside. This is where Single sign-on (OAuth, SAML) earns its place and where it tends to cause trouble.
Sound familiar?
The Single sign-on (OAuth, SAML) problems that bring people to us, and what you have once each one is fixed.
Usually a rotated signing certificate, a mismatched entity ID, or XML canonicalisation that the two sides disagree on. The browser trace tells you which.
Assertions verified and accepted
The IdP is sending http://schemas.xmlsoap.org/ws/2005/05/identity/claims/emailaddress and the app wants email. Users log in and land as the wrong person or get a 500.
Claims mapped to the app's schema
The registered callback URL does not match exactly, often a trailing slash or http versus https. OIDC is strict on purpose here.
Redirects allowed and verified
The offline_access scope was never requested, or the IdP policy does not grant it, so the app has no way to extend the session.
Refresh flow working end to end
The IdP pushes an operation the app does not understand, commonly a group membership change with a path the server rejects. Users stop appearing or stop being deactivated.
Users and groups sync again
The enterprise application exists but the user or their group has not been assigned to it, so Entra ID refuses the token request before the app ever sees it.
Assignments and roles in order
Server time has drifted beyond the five-minute window the assertion allows, usually on a VM without NTP. The failure looks random until you check the clocks.
Time synced, drift alerted on
An in-house or legacy application still uses its own user table and nobody wants to leave it that way before the next audit.
SAML and OIDC in front of it
What we do
Call us for one failing assertion, or hand us the whole identity project.
How it works
Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.
We fix the cause, not the symptom. If something is down, getting it back up comes first.
Backups, monitoring, updates and access, so the same problem can't quietly come back.
You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.
Often found together
We work on the whole stack, so the problem can be wherever it actually is.
Yes. We add a SAML service provider to your application, build the per-customer configuration, and test the flow end to end against each major identity provider. You get a setup guide your customers can follow, and we stay on call for the first few customer rollouts so edge cases get handled quickly.
Almost always an attribute mapping mismatch. The identity provider is sending a claim with one name or format, and the application is looking for a different one. We read the real SAML response from the browser, compare it against what the app expects, and align the mapping on whichever side is easier to change.
Yes. We register the OAuth client, wire up the authorization code flow with PKCE, handle nonce and state correctly, and link the resulting identity to your existing user records. If you also want to accept enterprise SSO through the same path we can set that up alongside.
We implement the SCIM 2.0 endpoints your identity provider expects, test them against Okta, Entra ID and whichever others you need, and work through the small differences in how each IdP interprets the standard. Then we set up alerting so a provisioning failure is noticed on the day it happens.
Yes, and the important part is the plan. We keep both paths working during cutover, map existing accounts into the new identity provider by email or employee ID, and give you a rollback route in case something unexpected shows up in production.
Estimates are free. Describe the application and which identity providers you need to support, and you get a price before any work begins.
Free estimate
Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.
Your message is with a real person, and we'll get back to you soon.
Rather talk now? Call us: