Microsoft Just Validated Our Architecture

Microsoft just shipped Scout at Build.

Scout is an always-on personal agent built on OpenClaw — the same open-source agent runtime that powers Tula, our patient agent at RealActivity.

Think of Tula as your patients' Scout.

Microsoft's insight with Scout: open-source agent capability is only half the equation. The other half is governance. Identity, audit, policy enforcement, controls your compliance team can actually verify. Their own words: every agent operates under its own governed identity, with scoped credentials and audit-ready policy conformance.

That is exactly the architecture we built for the patient side.

Tula is the open-source patient agent runtime (Apache 2.0, on OpenClaw). Aria is the governance plane. HIPAA-aligned identity, full audit trail, escalation policy, hospital-controlled model routing.

That side of the equation has been ungoverned. Patients are already using ChatGPT and Copilot to interpret their labs, manage their medications, and understand their diagnoses. They have their own Scout. It just has no governance, no grounding in their real record, and no connection back to their care team.

The conversation is happening. It is happening outside your governance.

Tula connects to the patient's real clinical record via SMART-on-FHIR. Aria wraps it with the governance layer the hospital requires. The patient gets a real Scout. The hospital gets verifiable control. Same open-source agent foundation Microsoft just validated in front of the whole industry.

If you are a CMIO or CIO watching Build this week and wondering what this means for your patients, I would like to have that conversation.