The New Car Dealership sector in the DecisioQ Decision Catalog is identified by AUTO-NCD. It currently covers 100 decision types across sales, CRM, inventory, F&I, customer experience, compliance, EV and connected services, facilities, strategy, and performance management.

DecisioQ integrations should load sector, category, decision, Decision Preparation Model, and scenario metadata from the Decision Catalog. Do not hard-code old template names, static decision lists, or legacy request models.

Recommended first deployment: start with a measurable advisory workflow such as internet lead prioritization, sales consultant assignment, dealer trade approval, trade-in offer review, lender selection, protection package recommendation, or aging inventory action.

New car dealership client applications should use the same discovery flow as the Business Decision Studio:

1. Load GET /decisioncatalog through the approved server-side Decision Catalog gateway.
2. Populate Sector from catalog.sectors[] and select AUTO-NCD for New Car Dealership.
3. Populate Category from selectedSector.categories[]. Example categories include Compliance & Risk, Customer Experience, CRM & Lead Management, EV & Mobility, Finance & Insurance, Inventory Management, Facility Operations, Digital Services, and Business Strategy.
4. Populate Decision from selectedCategory.decisions[] when available, or call /decisioncatalog/sectors/{sectorCode}/categories/{categoryCode}/decisions.
5. Display Decision Name only in the dropdown while keeping decisionId as the option value.
6. Load full decision detail from /decisioncatalog/decisions/{decisionId} before rendering Guided Data Collection.

Decision detail is the source of truth for criterion names, data guidance, scale, direction, profile options, scenarios, and overview metadata.

For regional deployments, keep DecisioQ technical identifiers stable and localize only user-facing labels and guidance. Unless the selected endpoint and decision schema explicitly support unit-qualified input, convert mileage, distance, pressure, temperature, mass, volume, speed, fuel consumption, and other measurements to the required canonical units before submission. Do not infer units or currency from locale.

Dealership integrations should map CRM, DMS, inventory, desking, F&I, service, OEM program, and customer-experience data into decision-specific Business Data. The selected decision detail defines which criteria are required and how each value should be measured.

Do not invent criterion IDs. Use the canonical criteria returned by /decisioncatalog/decisions/{decisionId}, including criterion overview metadata, unit, direction, scale, and data guidance.

Production integrations should keep credentials and bearer tokens on trusted servers or server-side gateways. Browser pages should use same-origin proxy handlers or their own backend service.

1. Trusted backend obtains a JWT token from identity.vinquery.com.
2. Application loads catalog and decision detail metadata.
3. User or system prepares Business Data for candidate leads, vehicles, deals, F&I options, offers, complaints, inventory actions, or strategic projects.
4. Call POST /api/v1/validate to validate and prepare the decision input before execution.
5. Call POST /api/v1/decide to return the authoritative Decision Result.
6. Display the recommendation, ranked alternatives, warnings, assumptions, and Explanation of Decision Result.
7. Keep execution trace and diagnostics available to integrators, but collapsed or hidden by default for business users.

Use /api/v1/decide for both Prepared Criteria Mode and business-data integrations that submit a businessData object. New public onboarding should prefer Preview Decision and Execute Decision terminology.

The exact fields depend on the selected New Car Dealership decision. The example below shows the integration style rather than a guaranteed schema for every decision.

{
  "decisionId": "AUTO-NCD-004",
  "decisionPreparationModelId": "AUTO-NCD-PROFILE-STANDARD",
  "scenarioId": "AUTO-NCD-SCENARIO-INTERNET-LEAD-PRIORITY",
  "businessData": {
    "candidates": [
      {
        "optionId": "LEAD-48192",
        "name": "Retail customer interested in 2026 SUV",
        "values": {
          "engagementScore": 91,
          "purchaseIntentScore": 84,
          "responseAgeMinutes": 18,
          "vehicleMatchScore": 88,
          "appointmentLikelihoodScore": 79,
          "grossOpportunityScore": 73
        }
      },
      {
        "optionId": "LEAD-48207",
        "name": "Lease renewal customer interested in EV",
        "values": {
          "engagementScore": 86,
          "purchaseIntentScore": 89,
          "responseAgeMinutes": 42,
          "vehicleMatchScore": 82,
          "appointmentLikelihoodScore": 84,
          "grossOpportunityScore": 68
        }
      }
    ]
  },
  "requestContext": {
    "correlationId": "dealership-crm-20260717-001"
  }
}

This pattern belongs in a trusted backend, not directly in browser JavaScript.

async function executeNewCarDealershipDecision(decisionInput, jwtToken) {
  const response = await fetch(`${process.env.DECISIOQ_BASE_URL}/api/v1/decide`, {
    method: "POST",
    headers: {
      "Authorization": `Bearer ${jwtToken}`,
      "Content-Type": "application/json"
    },
    body: JSON.stringify(decisionInput)
  });

  if (!response.ok) {
    throw new Error(`DecisioQ execution failed: ${response.status} ${await response.text()}`);
  }

  return await response.json();
}

• Store identity credentials, Decision API base URL, and proxy configuration in trusted server configuration or a secret manager.
• Do not expose bearer tokens or integration credentials in browser code.
• Use catalog metadata rather than hard-coded categories, decisions, criteria, profiles, or scenarios.
• Validate Business Data before execution and show user-friendly validation messages.
• Persist correlation ID, lead ID, customer ID, deal ID, VIN, stock number, decision ID, profile ID, scenario ID, selected option, score, and explanation for audit.
• Retain the successful decision response requestId with the host business transaction, then call POST /api/v1/decision-outcomes when the action and realized result are known.
• Require an overrideReason for overridden recommendations; never infer acceptance merely because a result was displayed.
• Monitor GET /api/v1/decision-outcomes/summary for acceptance rate, override rate, and realized-value aggregates. Treat outcome feedback as reporting evidence, not automatic model training.
• Redact customer identifiers, credit-sensitive data, lender details, deal notes, complaint notes, and sensitive margin information in logs and diagnostics.
• Monitor catalog load failures, token refresh failures, API latency, validation failures, account verification failures, and unusual exclusion rates.
• Define fallback behavior so CRM, sales, inventory, F&I, and customer-service workflows remain usable when DecisioQ is unavailable.
• Review Decision Preparation Models and scenario assumptions with sales, F&I, inventory, compliance, and management owners when dealership policy or OEM programs change.