Category boundary
Proof is useful. Interpretation is the next problem.
Provenance and authenticity systems can establish important facts about an input: where it came from, who signed it, whether it was transformed, or whether its origin can be verified. WITNESS is designed for the downstream question: what does the available evidence actually support?
| Layer | Primary question | Typical output |
|---|---|---|
| Identity | Who is this? | Identity or account assertion. |
| Provenance | Where did this come from? | Origin, signature, attribution, chain of custody, or authenticity metadata. |
| Analytics | What happened? | Measurements, trends, segments, and events. |
| WITNESS | What does the evidence support? | An evidence-backed Finding with alternatives, uncertainty, and traceability. |
Verified evidence can still be misunderstood.
Knowing that a signal is authentic does not automatically tell you how to interpret it. A verified metric can still be incomplete. Two legitimate sources can disagree. A real observation can support several competing explanations. Provenance improves the quality of the evidence layer; WITNESS investigates the meaning of that evidence.
Signal Integrity
Source authenticity is one dimension of trust, not the entire trust decision.
WITNESS Signal Integrity can consider provenance alongside structural patterns, contradictions, missingness, timing behavior, and other evidence-quality indicators before downstream reasoning consumes the signal.
No false equivalence