OutcomeReconIndependent settlement control

Proof / Recon Engine v1.1

QUALIFIED ENOUGH FOR A MANAGED PILOT. NOT PRESENTED AS ENTERPRISE PRODUCTION.

The benchmark is an environment-specific synthetic cold-start qualification designed to test exactness, Claim Graph deduplication and scale behavior. It is not a client case and not a universal hardware guarantee.

Synthetic qualification result

500,000synthetic outcomes
5,010,000synthetic events
20,000injected conflict outcomes detected out of 20,000
$19,800.00exact synthetic recovery

40,000 overlapping monetizable findings were collapsed by the Claim Graph into 20,000 unique economic claims, preventing double summation of the same economic dollar.

Recorded phase timing

PhaseRecorded time
Ingestion34.055 s
Index build3.395 s
Rules + Claim Graph21.084 s
Full exports23.474 s
Total82.012 s
Reference run: 100,000 outcomes + 1,002,000 events; 4,000/4,000 injected conflicts; $3,960 exact; 15.809 seconds.

Functional proof

Recon Engine v1.1 reports version 1.1.0 and passed 35 discoverable tests in the recorded release verification. The engine uses integer cents / micro-rates, versioned authority, source hashing, quarantine/failure records, multi-currency isolation and separate financial-versus-workflow states.

What this does not prove

  • No real client dataset has yet been qualified.
  • 10M+ rows have not yet been qualified.
  • No claim is made that the recorded timing will reproduce on every hardware or dataset.
  • There is no hosted multi-tenant SaaS, live connector layer, RBAC/SSO/SAML or SOC 2 / ISO 27001 certification in v1.1.
  • Intercom Fin is the only packaged public-policy profile currently functionally and scale-qualified; signed customer terms still override the public profile.