A conversation.
A checkable identity.

Bring the check into the workflow.

Remote support and business processes already happen through messages. Asking someone to leave that conversation, photograph a document and email it back creates more handling, more copies and another disconnected decision.

We are exploring a different interaction through the European Digital Identity wallet ecosystem: an operator sends a verification request, the person reviews it in their wallet, and the result returns to the operator's console.

Ask for evidence, not a blanket permission.

Our current proof of concept focuses on a name check using test credentials. The operator can provide an expected name, and the holder chooses whether to share the requested information. The request can expire, be cancelled or be declined.

The result supports a business decision. It does not itself authorise a payment, change a customer record or grant an AI agent permission to act. Those remain separate decisions with their own policy and approval requirements.

Identity is one part of safe AI enablement.

An AI-assisted workflow may gather information, prepare a request or help an operator understand a result. It still needs to distinguish the person, the evidence presented and the authority for the next action.

This is the same question behind Krnali SAR, Krnali Ops and Krnali VE. What is the work? Who can act? What evidence supports the decision? What can we verify afterwards? The messaging project explores the human identity side of that work; it is not a claim that all of these projects are already integrated.

A working proof of concept, with a clear test boundary.

The hosted implementation creates real requests through the official EUDI sandbox verifier. It uses test credentials, and successful presentation from an installed phone wallet is still awaiting acceptance testing. It is not a production identity verification service.