Principal Product Architect at Microsoft. I define identity and trust infrastructure for autonomous agents: how machines establish who they are, prove what they are authorized to do, and remain accountable when acting without a human in the loop.

The throughline in my work has always been trust and autonomy in networked systems. At Minutepay.fr, an early email-payments venture backed by Banque Directe, the problem was enabling people to transfer value safely online. At Planning Studio Inc., it was giving cities computational tools for modeling their own futures. At Microsoft, I helped ship the first version of Azure, scale Office 365, and advance open source when that position was still contested inside the company.

Now the object acquiring identity and autonomy is the machine itself. That shift — from people acting through software to software acting on their behalf — is the problem I work on today.

• Entrabot — A reference implementation that gives a device-local agent its own Microsoft Entra Agent ID and Agent User, with independent authentication, authorization, and attribution across Microsoft 365.
• Identity Research for Agent Management Using SPIFFE — A working prototype for governing agent-to-agent traffic with Entra Agent Identity, SPIFFE/SPIRE workload identity, and sidecar-enforced policy.
• Persona-Sati — A research program in persistent machine cognition: how an artificial persona maintains memory, identity, and agency across time and across agent bodies.

My work sits at the boundary of distributed systems, philosophy, and cognitive science. I also write at brandonwerner.com.

Serious technical conversations and collaborations: brandon@werner.ac