Feature · AI across MedIQ

Useful AI. Auditable AI. Production-ready AI.

MedIQ is AI-native — but we never bet patient outcomes on a single black box. Deterministic rules do the heavy lifting; governed AI adds judgment on top, fully audited and fully reversible.

How governed AI works Trust Center

De-identified prompts

We strip patient identifiers before any model call. Prompts log only a structured payload and an input hash.

Auditable decisions

Every AI call writes a row: input hash, decision, rationale, confidence — exportable for audit.

Deterministic-first

Rules run before AI. If a rule blocks, we never ship the prompt — and AI can never override a rule.

Managed by MedIQ

MedIQ centrally vets and assigns models per task. No tenant keys, no model selection, no metered surprise.

Encrypted secrets

Tenant configuration is encrypted at rest with AES-256-GCM and per-record IVs, gated to tenant admins.

Latency budget

Scrubber AI pass typically completes in under 1.5s p95 — fast enough for inline review during charge capture.

Scrubbing

Deterministic 60+ rule engine; AI risk pass adds context only humans usually catch.

Coding

15-stage pipeline over the national catalogs (ICD-10, CPT, HCPCS, NCCI, MUE) with governed AI refinement.

Appeals

Auto-drafted appeal letters with payer-shaped tone, citations, and tracked revisions.

Pipeline

A claim's trip through the AI layer.

Plain-English description of the path every claim takes — deterministic checks first, then AI, then a reversible patch your biller can accept or reject.

1
Ingest
Claim arrives from encounter or batch upload. Tenant + role checked.
2
Rule engine
60+ deterministic rules: crosswalks, modifiers, duplicates, payer rules.
3
De-identify
PHI stripped; structured payload + input hash prepared.
4
Model call
Governed AI returns a risk score and rationale — logged with input hash.
5
Patch
Suggested reversible change synthesized from rule + AI output.
6
Audit
Decision + confidence + input hash written to the audit log.

Want the governance brief?

We'll send the architecture overview, model review checklist, and audit-log schema.

Request the brief Governed AI →