I build PingKit because the network is where most everyday tech problems actually live: slow WiFi, a device you don't recognise, a connection that keeps dropping. The tools for understanding it were either enterprise-heavy, ad-riddled, or quietly shipping your data somewhere.
My background is in DevOps and systems engineering, keeping infrastructure reliable, observable, and secure. PingKit brings that same discipline to your home network: the diagnostics you'd expect from professional tooling, wrapped in an interface that explains what it finds in plain language. Everything that can run on your device does. There are no accounts, no ads, and no third-party analytics SDKs harvesting your activity. The only usage data collected is anonymous, aggregate counts of which features get used, with no identifier attached.
Why privacy-first isn't a marketing line here
Network data is some of the most revealing data there is. It describes every device you own and how you use them. That's exactly why PingKit keeps it on your device and syncs it only through your own private iCloud database, which even I can't see. Discovery and security scoring run on your device; the Security Scan also contacts a few internet services it needs, such as Shodan's InternetDB, which it asks what is visible on your own public address. So does device identification, most of the time. With Guardian, and once you have allowed PingKit's own AI service, a device the on-device classifier cannot name can be sent to that service with the technical details the device announces or answers with, such as its name, maker, open ports and service records. The same is true of AI answers: they are written on your device where Apple Intelligence is available, or with Guardian by Apple Private Cloud Compute, and PingKit's own service is the last resort. It asks your permission before it ever receives anything, and you can take that permission back. I would rather describe that exactly than round it down to "everything stays local", because that would not be true.