Loading…
Loading…
Scenarios are one of the clearest ways to evaluate how GlobiGuard behaves. Whether you start with guided examples, connected systems, or synthetic test data, the same review path stays visible: data enters, sensitive content is checked, policy decisions are applied, and results are recorded.
Preloaded examples that help teams understand the product flow quickly. These are best when you want to see the review path without connecting your own systems first.
A policy-review flow that detects sensitive fields and applies redaction before the model-facing step continues.
A financial-records review example showing how sensitive values are identified and handled before downstream processing.
An inbox-review example that classifies and prepares responses while keeping sensitive content out of the model request when policy requires it.
Connect supported systems and run the same review flow against your own records. This path is for teams moving from exploration into a more realistic product setup.
Connect Gmail or Outlook and inspect how sensitive content is detected and handled across real mailbox content.
Connect CRM records and review how contacts, notes, and related fields are screened before the model-facing step.
Use realistic but non-production data to test product behavior, policies, and review posture without moving customer records through the flow.
Generated healthcare-style records that make it easier to test health-data review policies without using real patient data.
Synthetic finance-oriented data that helps teams test sensitive-field handling before production rollout.
When you run a scenario, you can inspect:
Every supported scenario follows the same five-stage review path. This is the shared product model behind guided examples, connected data, and synthetic test data:
Raw data is ingested from the connected source (email body, CRM record, financial document).
GlobiGuard's detection engine scans every field for personal identifiable information using NER, regex patterns, and contextual analysis.
Detected PII is evaluated against your active policies. Data may be allowed, redacted, or blocked based on sensitivity level and context.
Sanitized data, with only the context allowed by policy, is sent to the configured provider step for summarization, classification, or other supported processing.
The result is returned and the supported review path records the detections, decisions, and outputs that matter for later inspection.
If the scenario API is enabled for your environment, list available scenarios with GET /api/demo/scenarios. Execute a scenario with POST to /api/demo/scenarios/:id/run. The response includes stage results, detections, policy decisions, and the final model output for that run.