00Decide — how the gate works

The verdict is not a guess.
It is the projection of facts onto rules.

Pure LLMs cannot be a payer's medical director — the same case produces three different denials across three runs. Pure rule engines cannot either — they cannot read a free-text note. Verum extracts probabilistically and decides deterministically. Below: the gate, made visible.

01Interactive — drag the dots

Rule projection, live.

Three threshold rules. Three clinical facts. As you move each fact across its threshold, the verdict on the right re-derives — in pure code, the moment your finger lifts. This is how the real engine works. No model is called.

INPUT · clinical facts
Drag the dots. Cross the threshold rule. Watch the verdict re-derive in pure code.
FACT · 01
Conservative therapy
8.0weeks
6
met
Mandatory · ≥ 6 weeks PT or equivalent
FACT · 02
Oswestry Disability Index
42score
30
met
Mandatory · ≥ 30 score
FACT · 03
Red-flag radiculopathy / neuro deficit
Rule · presence bypasses conservative-therapy gate (medical-necessity exception)
Absent
Try a preset
OUTPUT · deterministic verdict
No LLM. No regeneration. Same facts always produce the same answer.
VERDICT
APPROVE
Confidence
0.94
Rationale
"All mandatory criteria met."
Rule projection
  • conservativeTx ≥ 6w8.0w
  • odi ≥ 3042
  • redFlag bypassabsent
02A vs B — same input, two engines

Three runs. Three verdicts.
Or one.

IDENTICAL · input case
Same clinical note. Same policy. Run repeatedly.

"62F with 8 weeks of PT-documented chronic low back pain. ODI 42. Conservative trial of NSAIDs and structured PT failed. No red-flag findings on exam. Requesting lumbar MRI without contrast."

CPT
72148
ICD
M54.50
Payer
MCR
APPROACH A · pure LLM
⚠ NON-DETERMINISTIC
Run 1 of 3 — same input, different verdict each time.
VERDICT
PEND
confidence
0.68

"Conservative therapy possibly insufficient. Recommend additional records review."

APPROACH B · Verum
✓ DETERMINISTIC
Same input. Every run. Forever.
VERDICT
APPROVE
confidence
0.94

"All mandatory criteria met. PT ≥ 6w (✓ 8w). ODI ≥ 30 (✓ 42). No red flags required for approval."

Why it never changes
  • · Rules evaluated in pure code
  • · No temperature, no sampling
  • · Same facts → same answer, always
03Borderline tolerance · the ε band

A rule says ≥ 6 weeks. The fact says 5.5.

A rigid gate would deny. A clinically-aware gate routes to PEND@0.55 — "additional records needed" — when the fact is within ε of the threshold.

ε = max(2, threshold × 0.25) — calibrated so the borderline band scales with the threshold magnitude. Hard fails still dominate; only numeric mandatory rules participate.

Restores the legacy LLM "−0.15 borderline" tier, deterministically.

NUMERIC RULE · conservativeTx ≥ 6 weeks
0w
2w
4w
6w
8w
10w
12w
2w · DENY
5w · PEND
8w · APPROVE
< threshold − ε · DENY
within ε · PEND@0.55
≥ threshold · APPROVE
04Audit anatomy

Every verdict is six things.
Override any one. It re-derives.

01

Rule identity

Every gate references a stable ruleId — e.g. LCD-L34220-3.1. The verdict ties to it; the rule's text can change without the verdict losing its anchor.

ruleId: string
02

Criterion

The rule's machine-readable threshold — ≥ 6 weeks, present/absent, ICD prefix match. Same shape across every payer ontology so cross-payer comparison works.

criterion: { kind, threshold, ... }
03

Verbatim quote

The policy text the rule was encoded from — copy-paste from Aetna CPB 0236, CMS LCD L34220, or Carelon imaging. No paraphrase. Source of authority.

evidenceQuote: string
04

Met / unmet / borderline

The three states a criterion can produce. Borderline only applies to numeric rules within ε of threshold. Hard fails dominate aggregation.

status: "met" | "borderline" | "unmet"
05

Confidence

A separate scalar from the fact extraction. Surface-able as a calibration metric in evals. Borderline verdicts cap at 0.55.

confidence: 0..1
06

Operator hand

A medical director can promote any fact or rule outcome. The override is itself audited — original verdict + override + new verdict, all retained.

DecisionOverride · ResolutionAudit
05BRIDGE · image evidence into the gate

The report hedged.
Verum reads the image.

Dictated radiology reports hedge. “L5-S1 narrowing on imaging, etiology unclear, recommend clinical correlation.” The gate pends on the documentation gap — surgical decompression authorization requires imaging-confirmed moderate-to-severe stenosis with documented etiology. Patient round-trips for a clarifying read; case takes another week.

BRIDGE closes that loop. Inside /review, the UM reviewer clicks 📷 Read this image with Verum · Vision. Findings stream into a right-side drawer in 3-8 seconds. The reviewer attests the relevant ones and clicks Promote. The deterministic gate re-derives against the augmented fact bank — the BRIDGE-extended rules now fire MET — and the verdict flips PEND → APPROVE.

The downstream artifacts inherit the evidence. The appeal letter generator and the SOAP note generator both cite the AI-attested image findings inline, with full attribution — reviewer name, vision-language model, timestamp, verbatim visual evidence quote. The decision-packet PDF gets a Page-1 banner and a Page-3 audit section. Append-only on the Decision: every promotion is reversible via the existing revert chain.

BRIDGE-extended rules carry a brass “BRIDGE” pill in the cited-rules list — so a reviewer can always tell which rules derive from core Medicare / CMS policy versus the additive AI-evidence ontology layer.

BRIDGE · 3-step flow
  1. 01
    READPEND

    Verum · Vision reads the attached image

    MedGemma 4B foundation, fine-tuned for our 4 imaging modalities. 3-8 seconds warm. Emits structured findings keyed to the policy vocabulary.

  2. 02
    ATTESTPEND

    Reviewer attests + promotes

    High-confidence findings pre-checked; reviewer toggles + clicks Promote. Audit captures the reviewer’s name, timestamp, model, verbatim visual evidence.

  3. 03
    RE-DERIVEAPPROVE

    Gate re-evaluates · verdict flips

    BRIDGE-extended rules fire MET against the augmented fact bank. Deterministic gate re-derives in pure code. Appeal letter + SOAP note auto-cite the AI evidence with attribution.

sample · bridge-lumbar-decompression-imaging-gap

Open the workspace

Run a real case. See the gate work.