Traffic Analysis
Learning from the pattern of encrypted communication - who talks to whom, when, and how much - without reading any of the content.
Traffic analysis extracts information from the shape of communication rather than its contents: which endpoints are talking, at what times, in what volumes, and with what regularity. Encryption protects the payload and leaves most of that shape exposed, which is why it is a discipline in its own right - the metadata is often sufficient, and it is available immediately rather than after the cryptography is broken.
For a data subscriber the exposure is specific and underrated. The set of streams a firm subscribes to is a map of its interests, and the moment a new subscription starts is frequently more informative than any record inside it. Volume and query rhythm describe how a desk operates; a burst of interest in one subject area describes what it is building. This is the metadata half of the same problem harvest now, decrypt later poses for content, and it is why the shape of a connection - one long-lived stream rather than thousands of targeted requests - is itself a privacy property.
Real-time data, at edge speed.
A live evaluation measures EdgeOrigin against your coverage requirements: decision-ready real-time data, delivery latency into your systems, and record-level provenance.