The access logs were blank, and that was the point.

Building an anonymous analytics system inside a VPC takes more than toggling a privacy flag. To keep data truly private, every request has to bypass public exposure, every packet must remain inside secured infrastructure, and every metric must still be delivered with precision. This is where a private subnet proxy deployment becomes the backbone of the setup.

Start with a VPC that contains no direct connections to the internet. The analytics ingestion endpoint should live inside a private subnet. A proxy then becomes the controlled escape route, channeling requests through internal paths and out via tightly managed NAT gateways or VPC endpoints. This structure enforces zero direct exposure, preventing origin IP tracing and blocking any route outside your control.

A proper deployment routes analytics events through the proxy with TLS termination inside the private network. From there, sanitized logs can pass into your processing layer. By pushing data through VPC logs, CloudWatch metrics, and internal load balancers, you ensure observability without sacrificing anonymity. Query systems and storage remain isolated, accessible only through the proxy path.

Scaling this configuration means automating proxy nodes in multiple availability zones, balancing load internally, and locking down security groups to explicit inbound sources. Every new service or pipeline should connect only through the private subnet mesh, ensuring data never leaves the bubble unguarded.

The advantage of this design is control. No hidden calls, no shadow APIs, no metadata leakage. Just raw, anonymous analytics data processed in a hardened network segment.

You can see it live in minutes with hoop.dev — deploy a full anonymous analytics VPC private subnet proxy without writing endless configs, and watch privacy and performance run side by side from the start.