Techbanq

Software · Nafath integration

Nafath verifies the person; your application still governs the session

The national single sign-on answers one question with authority: this is the person. Everything after that answer is yours to engineer — which actions demand a higher assurance level, how long a session lives before the question is asked again, what a user can do when the platform is briefly unreachable, and how consent and identity data are handled under PDPL.

Workstation showing terminal output during a security review

Serving Saudi Arabia from our Riyadh office and our engineering hubs in Lahore and Karachi

Enterprise delivery credentials

  • ISO

    ISO/IEC 27001

    ISO/IEC 27001 certified

Certifications & registrations

The sign-in path

How a Nafath sign-in travels from your login screen to a session you govern

The flow is standard OpenID Connect with a national identity behind it. The drawing exists to show where the platform’s responsibility ends and yours begins — earlier than most teams assume.

A flow: the application redirects to Nafath with a scoped request, the user approves on their device, the backend exchanges and validates the tokens, and a session is issued under the application’s own policy — with a branch for a planned fallback when the platform is unreachable.

  1. 01

    Redirect with intent

    Your application sends the user to Nafath carrying exactly the request the action needs — no broader scope, no higher assurance level than the operation justifies.

  2. 02

    Approval on the user’s device

    The person confirms the request in their Nafath app. Your application waits on a state it can time out and explain, not on a spinner that hangs forever.

  3. 03

    Code exchange and validation

    Your backend exchanges the code, validates the tokens and signatures against the platform’s published keys, and checks that the assurance level returned matches what the action demanded.

  4. 04

    A session on your terms

    Nafath verified the person; the session is yours. Its length, its refresh rules and the moments that force re-verification are application policy, decided by the risk of each action.

Always available

When Nafath is unreachable

A planned degraded path — a queued action, a retry with honest messaging, or an alternative authentication route for the operations that permit one. What the fallback must never be is a silent bypass of identity for an action that required it.

The decisions Nafath leaves you

What a verified identity from Nafath does not decide for you

Nafath supports more than one level of identity assurance, and the level an action needs is a judgement about that action, not a platform default. Viewing a statement, changing a beneficiary and authorising a high-value operation are three different risks, and a well-designed integration asks for the stronger verification only where the risk earns it — because demanding the maximum everywhere trains users to approve prompts without reading them.

The same is true of sessions. The platform authenticates a moment; your application decides how long that moment is trusted. A session policy is a set of named decisions — idle timeout, absolute lifetime, which operations force a step-up, what a refresh may extend — and writing them down before the build is what separates an integration from a login button. When those decisions are left to framework defaults, the defaults become your security policy without anyone having chosen it.

Identity data under PDPL

Aligning a Nafath integration with PDPL before the first real sign-in

A national identity flow moves personal data by definition, so the data-protection decisions are part of the integration design, not a review that happens after it ships.

Consent that means something

The user is told, at the point of sign-in, what is requested and why, in language that matches what actually happens — and the consent is recorded with the request it covered.

What that protects: A consent trail you can produce, not reconstruct.

Minimisation of what is kept

The application stores the identifier and attributes the service genuinely needs, not everything the flow returned — and the design names which fields those are.

What that protects: Less personal data at risk than the flow could have collected.

Retention with an end

Identity attributes carry a retention decision made when they are stored, so data does not accumulate simply because deleting it was nobody’s job.

What that protects: A data estate that shrinks on schedule instead of growing by default.

Access that leaves a trace

Reads of identity data inside your own systems are logged and attributable, because the obligation does not end when the data arrives from the platform.

What that protects: An answer to “who looked at this record” that does not depend on memory.

Sandbox first

Proving the Nafath flow in the sandbox before a real user meets it

The happy path is the easy part to test. The rehearsal that matters is the unhappy one, because a national sign-in that fails confusingly costs you users who cannot tell whose fault it was.

STAGE 01

Build against the sandbox, not the docs alone

The flow is implemented and exercised end to end in the platform’s test environment, so behaviour that documentation summarises is observed rather than assumed. An implementation matched to observed behaviour.

STAGE 02

Rehearse denial, timeout and abandonment

A user who declines the prompt, a request that expires, a person who walks away mid-flow — each path ends in a state your interface explains, not a stack trace or a frozen screen. Failure states a support desk can name.

STAGE 03

Exercise token validation and key rotation

Signature checks, expiry handling and the platform rotating its keys are tested deliberately, because these are the failures that otherwise arrive in production at the worst hour. A validation layer that survives routine platform change.

STAGE 04

Prepare the onboarding your organisation will make

Production access is granted by the platform operator to your organisation, not to us. We assemble the technical material that request needs and make sure the build already meets the specifications it will be judged against. An application ready for the operator’s process.

Our role

Where our Nafath engineering ends and the operator’s onboarding begins

  1. 01

    The flow, engineered properly

    We build the OpenID Connect integration, the token validation and the error handling against the platform’s published specifications, with the unhappy paths designed rather than discovered.

  2. 02

    The policy, made explicit

    Assurance levels per action, session lifetimes, step-up moments and the fallback route are written down as decisions you approved, so the security posture is chosen rather than inherited from defaults.

  3. 03

    The data handling, PDPL-aligned

    Consent capture, minimisation, retention and access logging are built into the flow from the first design, because bolting them on later means re-plumbing the very paths the data took.

Onboarding, credentials and production approval are granted by the platform operator to your organisation; we prepare you for that process, build against the specifications, and hand over an integration your own team can operate and extend.

Common questions

What buyers ask

Do you get our application approved to use Nafath?+

No, and no vendor honestly can. Onboarding to Nafath — the credentials, the approvals and any agreements — is a process between the platform operator and your organisation. What we do is make that process as short as your side can make it: we build the integration against the published specifications, prove it in the sandbox including the failure paths, and prepare the technical material the onboarding asks for. The distinction matters, because a supplier who implies they can grant you access is describing a relationship with the operator that is yours, not theirs.

Should Nafath be the only way to sign in to our application?+

It depends on who your users are and what your actions risk. For services whose users all hold Saudi digital identities and whose operations genuinely need verified identity, Nafath as the sole route is a clean design. But many applications serve mixed audiences, or have low-risk actions a lighter authentication could carry, and every application needs an answer for the minutes when the platform is unreachable. We design that answer explicitly: which operations permit an alternative route, which must wait, and how the interface explains the difference — because the improvised version of that decision, made during an outage, is always worse.

What does integrating Nafath change about how we handle sessions?+

It sharpens a responsibility you already had. Nafath authenticates a moment; your application decides how long to trust it, and a verified national identity raises the stakes of that decision rather than removing it. The session policy — idle timeout, absolute lifetime, which operations force re-verification at a stronger assurance level — should be written per class of action, because a banking-grade operation and a profile view do not deserve the same trust window. We treat that policy as a named deliverable of the integration, reviewed with you, so it reflects your risk appetite instead of a framework default nobody chose.

Next step

Design your sign-in with us

A design session that maps which of your actions need Nafath, at what assurance level, and what the session and fallback policy will be — before credentials are even requested.