Skip to content
Simulation · not operational100% synthetic data · the demos’ human decisions were simulated in batch
planlang
ESEN

Demo A · run suscripcion-planlang-a-002-200-v1.5 · one real trace

One case, end to end

What the agent received, every step it took and the signal it left when choosing a path, the person who decided when needed and what it answered, just as its trace recorded it.

This run’s human decisions were simulated. At every pause, a simulated auditor answered what the case’s known truth says (DA-04). Cases, members and the benefits plan are synthetic.

The case at a glance

Read as

Reading as leader. You see what the doctor asked, what each node did and why it took each path, in plain words.

Reading as expert. Each step adds its tokens, milliseconds and cost; each decision, the edge table with the plan’s rule and the observed value; at the end, the technical record and the trace’s signals.

A-126missing: 3 clarifications

Approvedwith a personmatches the known truth
What should happen

The doctor needs three replies: if the clarification limit is below three, a human auditor decides and approves because doubt favours the member.

What happened

Final decision: approve, with a person.

Takes in

  • The doctor’s text
    Referral note. 30-year-old woman. Symptoms for several months that have not improved with conservative care. Request: Exercise stress test. Thank you.
  • The attached orderSYN-P-008 · care ambulatoria · Order signed by the treating doctor.
  • Member: woman, 30 years oldname, ID, medical record, phone and email are masked before the model (decision D1)

Does

10 steps
enrutadorextractoraclaracionextractoraclaracionextractoraclaracionpausa_humanaredactorguardia_salida

The doctor requested “Exercise stress test” for a 30-year-old woman. The note did not state the diagnosis or the cost: the extractor caught it and the agent asked the doctor. After 2 questions the data were still incomplete. The doctor did not complete the data within the clarification cap U3: the case went to an auditor, who decided to approve. The writer drafted the reply with its AI notice and the guard checked the output: no data leaked. It took 40.2 s.

Delivers

  • The reply to the member, with its AI notice
  • The trace, with its fingerprintcdc5f5bdd2d3…
40.2 send to end
10steps
6model calls
17,577tokens
0.0650nominal US$, not billed

Expert

The trace’s technical record

Trace
planlang-trace/v1 · cdc5f5bdd2d3cf2df228481ca6660d24287629ae8ad7729570d1f6930ed312b5
Run
suscripcion-planlang-a-002-200-v1.5 · variant multiagente · result completo
Errors
provider: none · schema on handoff: no
Known truth
decision aprobar · must escalate yes · cycles needed 3
Case
faltante · faltante_tres_ciclos · seed planlang-a-002

The path, step by step

real · steps and edge decisions

Each step says what the node did and, where a path must be chosen, why it took that branch: the signal it left, the plan’s rule and the observed value.

  1. 1
    enrutadorno model

    Classified the care type: outpatient.

    type enrutador · in 0 · out 0 tokens · 1 ms · US$ 0.000000 · schema retries 0

    neither an emergency nor an exempt service: on to the extractor

    1Signaltipo_atencionRule= urgenciaObservedambulatoriaHolds?noBranch—
    2Signalservicio_exentoRule= trueObservednoHolds?noBranchextractor
  2. 2
    extractor6.2 s · 4,305 tokens

    Read the doctor’s text and turned it into fields, with a stated confidence.

    type modelo · in 4019 · out 286 tokens · 6224 ms · US$ 0.015138 · schema retries 0

    data are missing: it asks the doctor

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalcampos_faltantes_countRule> 0Observed2Holds?yesBranchaclaracion
  3. 3
    aclaracion8.2 s · 1,289 tokens

    Asked the doctor (cycle 1).

    Question«Para continuar con la autorización de la prueba de esfuerzo, necesito que indique el diagnóstico principal (con código CIE-10, si lo tiene) que justifica la solicitud y el costo estimado del procedimiento.»

    Answer«Ver historia clínica adjunta.»

    In the original Spanish: the model asks in the case’s language.

    type modelo · in 1159 · out 130 tokens · 8247 ms · US$ 0.003986 · schema retries 0

    back to the extractor with the answer

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalciclos_aclaracionRule≥ U3 = 2Observed0Holds?noBranchextractor
  4. 4
    extractor6.6 s · 4,340 tokens

    Read the case again with the doctor’s answer.

    type modelo · in 4033 · out 307 tokens · 6602 ms · US$ 0.015385 · schema retries 0

    data are missing: it asks the doctor

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalcampos_faltantes_countRule> 0Observed2Holds?yesBranchaclaracion
  5. 5
    aclaracion6.5 s · 1,285 tokens

    Asked the doctor (cycle 2).

    Question«Para continuar con la solicitud de prueba de esfuerzo, ¿podría indicarnos el diagnóstico (o sospecha diagnóstica) que justifica el estudio y el costo estimado de la prueba?»

    Answer«Aclaración del médico. Diagnóstico: Enfermedad coronaria estable.»

    In the original Spanish: the model asks in the case’s language.

    type modelo · in 1158 · out 127 tokens · 6473 ms · US$ 0.005016 · schema retries 0

    back to the extractor with the answer

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalciclos_aclaracionRule≥ U3 = 2Observed1Holds?noBranchextractor
  6. 6
    extractor7.3 s · 4,773 tokens

    Read the case again with the doctor’s answer.

    type modelo · in 4074 · out 699 tokens · 7257 ms · US$ 0.019539 · schema retries 0

    data are missing: it asks the doctor

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalcampos_faltantes_countRule> 0Observed1Holds?yesBranchaclaracion
  7. 7
    aclaracionno model

    Asked the doctor (cycle 3).

    type modelo · in 0 · out 0 tokens · 0 ms · US$ 0.000000 · schema retries 0

    it reached the clarification cap: on to a person

    1Signalproveedor_no_disponibleRule= trueObservednoHolds?noBranch—
    2Signalciclos_aclaracionRule≥ U3 = 2Observed2Holds?yesBranchpausa_humana
  8. 8
    pausa_humanano model

    An auditor saw the full case and decided: approve.

    type pausa_humana · in 0 · out 0 tokens · 0 ms · US$ 0.000000 · schema retries 0

  9. 9
    redactor5.4 s · 1,585 tokens

    Wrote the reply to the member, with the AI notice.

    type modelo · in 1344 · out 241 tokens · 5415 ms · US$ 0.005931 · schema retries 0

  10. 10
    guardia_salidano model

    Checked the output: 0 findings, severity 0.

    type regla · in 0 · out 0 tokens · 0 ms · US$ 0.000000 · schema retries 0

The human pause: what the auditor saw

real · interrupt payload

No full denial goes out without a person. The agent stops and hands the auditor the full case: why it stopped, what it read, the original text, the evidence and the counter-evidence.

Why it stopped

The doctor did not complete the data within the clarification cap U3.

payload.motivo: Edge 2 of aclaracion: ciclos_aclaracion (2) at least umbral.U3 (2).

signal ciclos_aclaracion · declared threshold umbral.U3 · applied 2 · node pausa_humana · step 8 · role auditor

Evidence

  • Extractor confidence 0.7 below 0.75.
  • Details still missing (costo_estimado) after 2 clarifications.

Counter-evidence

  • Doubt about what a service covers is resolved in the member's favour (Law 1751, art. 8).

What the extractor read

costo_estimado — · diagnostico SYN-D-08 · procedimiento SYN-P-008 · urgencia no · confidence 0.70

The rest of the case it received

  • Order: SYN-P-008 · care ambulatoria · Order signed by the treating doctor.
  • Coverage: not yet checked when the agent stopped
  • Clarifications: 2, with their questions and answers
What the auditor answered

aprobar

Simulated auditor: in this batch run it follows the case’s known truth (policy sigue_verdad_conocida). In production a person decides with this same payload.

The reply and the guard

real · salida_final and guardia_salida
What the member receives

Your request for the stress test (SYN-P-008) has been approved. A person reviewed the request before making this decision. You can go ahead with the procedure as your provider instructs.

Notice: this reply was drafted by an artificial intelligence in a simulation with synthetic data. No denial is issued without review by a person.

Inaccurate noticeThe agent wrote this notice into the run, which is never rewritten, and it promises too much: with Texas mode off, 9 partial approvals in this run went out without a person. It holds for a full denial: from plan v1.5.1 on the agent writes it that way, and this run is from v1.5.

Output guard: fixed rules, no model

Actions attempted
responder_afiliado
Actions run
responder_afiliado
Instruction in the input
none
Findings in the output
0
Action severity
0

The 18 signals the trace leaves

Expert

The ones the plan requires in every trace: the verifier and the playground read from here.

campos_faltantes_count
1
carga_detectada
no
ciclos_aclaracion
2
contradiccion_orden_texto
—
costo_estimado
—
decision_final
aprobar
error_proveedor
—
latencia_total_s
40.22
modo_texas
no
nodos_visitados
enrutador → extractor → aclaracion → extractor → aclaracion → extractor → aclaracion → pausa_humana → redactor → guardia_salida
pausa_humana
yes
propuesta
—
proveedor_no_disponible
no
senal_confianza
0.70
servicio_exento
no
severidad_accion
0
tipo_atencion
ambulatoria
tokens
17577