Azure provides centralized governance, orchestration, and lifecycle management. Azure is the management layer — not the compute destination. All AI execution occurs locally.
Fleet management, device onboarding, lifecycle control, secure OTA updates, policy distribution, identity services, anonymized telemetry collection, API orchestration, operational visibility, and multi-tenant console.
Process raw sensor or operational data. Store PII, PHI, or classified content. Run AI inference. Access unprocessed telemetry. Serve as primary compute. Handle data that has not been hardware-anonymized.
Azure IoT Hub manages provisioning, registration, health monitoring, and decommissioning. Secure identity per node — Azure never accesses processed data.
Firmware, model, and configuration updates signed and distributed through Azure. Nodes verify signatures before applying. Automated rollback on failure.
Operators define compliance rules, workload constraints, access controls, and update schedules from a centralized Azure console.
Operational metrics — CPU, inference throughput, model version, health status — reported to Azure after hardware anonymization. No raw content.
Azure API Management provides secure gateway for enterprise integrations — connecting anonymized outputs to existing IT systems without exposing the edge data plane.