Every complex event leaves a signature.
AnnieGuard reconstructs it.

Glass-box technology, built for the unknown.

ORIGIN //

ORIGIN reconstructs the causal architecture of biological and chemical events from existing data—without requiring prior access to the causal agent, physical custody of a sample, or an established explanation.

Built for novel events. No event-specific training data required.

Deterministic. Traceable. Mechanistic.
Same inputs. Same reconstruction. Every time.

{See a Reconstruction}{>>}

CASCADE //

CASCADE reconstructs the causal architecture of disease from clinical and genomic data—tracing each case from its founding event through progression to therapeutic vulnerability.

Single-sample analysis. Existing clinical and genomic data.
No proprietary inputs. No new sequencing.

Deterministic. Traceable. Mechanistic.
Every causal claim is linked to its source data and assigned an explicit confidence rating.

{See a Reconstruction}{>>}

Causal Intelligence in Practice

{01}

Outbreak Attribution

ORIGIN demonstration.

Biological-event reconstruction and source attribution from population and surveillance data.

{02}

Defense & Intelligence

ORIGIN demonstration.

Biological and chemical threat reconstruction under operational time constraints.

{03}

Therapeutic Development

CASCADE demonstration.

Disease architecture reconstructed from clinical and genomic data to identify therapeutic vulnerabilities.

Where pattern recognition stops.

Conventional systems identify patterns. AnnieGuard reconstructs the causal architecture behind them from existing data. Deterministic, traceable, and mechanistic. Built for novel, high-consequence environments where auditability is non-negotiable.

{01}

DETERMINISTIC

Same inputs. Same reconstruction.
No stochastic drift. No hidden variability.

Reproducible, auditable, and reviewable across runs.

{02}

BUILT FOR THE UNSEEN

Novel events. Previously unseen cases.
No event-specific training data required.

One underlying methodology across biological and chemical domains.