PYRNOVAResearch
Research

An evidence-led approach, tested against the past.

Methods are examined the way conclusions are: with explicit evidence, point-in-time discipline and a stated way to be wrong.

Point-in-time reconstruction

Knowable then vs known now

To ask whether a consequence would have been knowable, the past has to be rebuilt without leaking the present into it. An AS OF view excludes later evidence and preserves the assessment state that existed at that moment.

t0AS OF cutoffnow AS OF eligible excluded from earlier view

Historical replay

Illustrative thesis lifecycle
2026-08-24OBSERVEDAcquisition strategy notice retained as source record.
2026-08-24ASSESSEDRoute to market shift derived; materiality High, confidence Moderate.
2026-09-02STRENGTHENEDSecond reference aligns with the consolidation direction.
2026-09-11MONITORINGVehicle access unresolved; prior assessment state preserved.

Validation discipline

Historical replay

Reasoning is exercised over retained historical events, so behaviour can be inspected rather than asserted.

Future exclusion

An AS OF view is reconstructed using only evidence available before its cutoff. Later evidence is excluded.

Falsification

Each conclusion records a falsifier, the evidence that would contradict it, so it can be challenged directly.

Provenance

Source identity, references and observation times are retained, so a claim can be traced to what it rests on.

On results

Pyrnova is early. This page describes the research approach and the disciplines applied to it. It does not claim predictive accuracy or a published benchmark result.

An evaluation tests Pyrnova against a real external change affecting your organization.

See the company-specific consequence, the evidence behind it, the uncertainty it keeps visible and the point in time it was knowable.

See how evaluation works