claim	evidence_class	environment	period	denominator_or_basis	boundary
204 UTC dates contain retained source records.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	204	Dates with records are not uninterrupted uptime or proof that every collector ran on every date.
At least 11 confirmed collector families are represented.	LOWER BOUND	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	11	A conservative lower bound, not a complete deployment census.
10 measured domains are confirmed.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	10	Domain presence does not establish equal coverage, independent sensors or customer outcomes.
351 selected files were retained and checked.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	351	A bounded selected population that includes copies, not all data ever produced by Raven.
229 byte-distinct contents were profiled.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	229	Different contents can still overlap. This is not a globally deduplicated event count.
4.86 GB of physical retained data was assessed, with copies included.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	351 selected files	Storage volume is not a count of unique measurements or a measure of system performance.
At least 1,949,874 distinct payloads were established within one dataset.	LOWER BOUND	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	1949874	A conservative single-dataset lower bound, not globally unique measurement events.
The earliest producer-bound measurement in this scope is 14 February 2026, 13:24:31 UTC.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	Earliest retained eligible measurement	Not a claim that Raven had no earlier measurements. File creation dates do not define measurement start.
501 comparable LTE/fiber acquisition cycles were recovered.	OBSERVED	Raven-owned environment only	8 to 10 June 2026	501	Comparison 1 and Comparison 2 share the same 501 acquisition cycles. Their counts are not additive. These are not independent incidents or a customer benchmark.
Comparison 1 fiber RTT: median 10.9795 ms, P95 11.09915 ms and maximum 18.046 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	500	Valid capture-level mean RTT observations in one same-destination ICMP comparison. Both comparisons share the same 501 cycles and are not additive. Not a universal fiber-performance claim.
Comparison 1 LTE RTT: median 33.447 ms, P95 67.394 ms and maximum 923.652 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	501	Valid capture-level mean RTT observations in the same comparison. Fiber and LTE have different valid denominators. Both comparisons share the same 501 cycles and are not additive.
Comparison 1 median paired LTE minus fiber RTT difference: 22.407 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	500	A paired timing difference, not an incident lead or business benefit. Both comparisons share the same 501 cycles and are not additive.
Comparison 2 fiber RTT: median 10.405 ms, P95 14.124 ms and maximum 14.381 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	501	Valid capture-level mean RTT observations in a separate same-destination ICMP comparison. Both comparisons share the same 501 cycles and are not additive. Not a universal fiber-performance claim.
Comparison 2 LTE RTT: median 30.865 ms, P95 77.114 ms and maximum 1015.363 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	501	Valid capture-level mean RTT observations in the same comparison. Both comparisons share the same 501 cycles and are not additive. Not a universal LTE-performance claim.
Comparison 2 median paired LTE minus fiber RTT difference: 20.379 ms.	DERIVED	Raven-owned environment only	8 to 10 June 2026	501	A paired timing difference, not an incident lead or business benefit. Both comparisons share the same 501 cycles and are not additive.
Measurements were recovered across a multi-host Raven-owned environment. Two hosts were directly audited and retained archives from a third host were assessed separately.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	Two directly audited hosts and retained third-host archives	Archived evidence and direct audit are separate. No device model or collector-to-device mapping is disclosed.
624 retained source validations supported 208 generated envelopes.	OBSERVED	Raven-owned environment only	8 to 9 June 2026	624 source validations; 208 envelopes	Scoped historical processing. Source files and envelopes are different units and are not added.
2,000 retained input events corresponded to 2,000 normalized outputs.	OBSERVED	Raven-owned environment only	30 May 2026	2,000 inputs; 2,000 outputs	Scoped normalization evidence does not establish production readiness or approved decisions.
24,482 collector-health evaluations and 121,522 Runtime-health evaluations were retained.	OBSERVED	Raven-owned environment only	9 June to 4 September 2026	24,482 collector-health evaluations; 121,522 Runtime-health evaluations	Separate periodic evaluation populations, not independent incidents, customer outcomes or decisions.
Reconstructed threshold chronology was observed, but no qualifying evidence-ready, recommendation or approved-decision lead was established.	NOT AVAILABLE	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	No qualifying downstream lead established	Threshold chronology is retrospective. No prediction, recommendation lead or approved-decision lead is claimed.
This release demonstrates real retained measurements, multiple hosts and domains, a bounded LTE/fiber comparison and scoped Runtime processing.	OBSERVED	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	Scoped evidence only	The release separates measured signal and processing evidence from unsupported decision and outcome claims.
This release does not establish customer validation, ORLEN or car-wash performance, production readiness, recommendation lead, approved decision lead, downtime reduction or ROI.	NOT AVAILABLE	Raven-owned environment only	1 February 2026 to 4 September 2026, 12:14 UTC	Not established in this scope	No business or customer-performance claim may be inferred from these measurements.