Scan Management

The Scans page is the operations center for all scanning activity. It provides a real-time view of active scans, a queue of pending jobs, full scan history, and controls to start, pause, resume, stop, and cancel scans.

Starting a New Scan

1. Navigate to Scans in the Customer Dashboard sidebar.
2. Click New Scan to open the scan configuration dialog.
3. Configure the scan parameters:

Connector Selection

Select the data source to scan from the dropdown. Only connectors with Active status are available. If no connectors are active, you will need to set one up first under Connectors.

Scan Types

For Targeted scans, an additional text field appears where you enter the path or prefix to scan (e.g., data/exports/2026/).

Classifier Template

Select a classifier template to control which classifiers run during the scan. If a default template exists, it is pre-selected automatically.

• Template selected — Only classifiers in that template are active for this scan
• None selected — All globally enabled classifiers run

See Classifiers for details on managing classifier sets.

Scan Profile

Control how intensively the scanner interacts with your data source:

4. Click Start Scan to submit the job.

Scan Lifecycle

Every scan progresses through defined states:

Created → Queued → Running → Completed
                     ↕           ↓
                   Paused     Stopped
                     ↓
                  Cancelled

Scan Controls

Active and paused scans expose four control actions:

Pause

Pauses a running scan at the current checkpoint. Workers finish processing their current file and then idle. The scan state is saved so it can be resumed later.

• Available when scan status is Running
• The scan transitions to Paused status
• Polling interval slows to 30 seconds while paused

Resume

Resumes a paused scan from its saved checkpoint. Workers pick up where they left off.

• Available when scan status is Paused
• The scan transitions back to Running status
• Polling interval returns to the active rate

Stop

Gracefully stops a scan and saves all partial results. Findings collected up to the point of stopping are preserved in the Data Catalog. The scan is marked as Stopped in history.

• Available when scan status is Running or Paused
• Partial findings are retained and visible in the Data Catalog
• The scan cannot be resumed after stopping

Cancel

Cancels a running scan. Workers finish processing their current resource, persist all findings discovered so far, and exit cleanly. The scan transitions to Cancelled status with an accurate coverage report.

• Available when scan status is Running or Paused
• Findings are preserved — all PII detected before cancellation is saved and visible in the Data Catalog
• The coverage report shows exactly which resources were scanned and which were not yet reached
• The scan cannot be resumed after cancelling

Pre-Scan Cost Estimate

Before a scan starts, the platform estimates the compute cost based on the number of resources, total data volume, and connector type. This estimate is displayed in the scan confirmation dialog.

If the estimated cost exceeds your configured ceiling, the scan is blocked with a clear message explaining the estimate and the ceiling. You can:

• Reduce scope — Narrow the scan to a specific prefix or path
• Switch to Light profile — Lower parallelism reduces cost
• Increase the ceiling — Ask your administrator to raise the per-scan cost ceiling in Settings

Scan Monitoring

KPI Summary

The top of the Scans page displays four key metrics:

• Total Scans — Lifetime count of all scans
• Active — Currently running scans
• Queued — Scans waiting for available workers
• Avg Duration — Average scan completion time

Active Operations

The Active Operations panel shows all currently running scans with:

• Connector name and provider — Which data source is being scanned
• Phase indicators — Visual progress through Discovering, Scanning, Analyzing, Complete
• Progress bar — Animated gradient bar showing percentage complete
• Current stage — Text description of the current pipeline phase
• PII Findings — Real-time count of findings detected so far
• ETA — Estimated time remaining

Concurrency Queue

Scans that are waiting for available workers appear in the queue panel with their queue position number. Scans execute in FIFO order within tier concurrency limits.

Scan History

The Scan History table shows all past scans with:

Status badges use the following color coding:

Scan Tier Limits

Scan capacity is governed by your subscription tier:

Scan API

Scans can also be triggered and managed via the API:

# Start a scan
POST /api/v1/connectors/:connector_id/scan
{
  "scan_type": "full",
  "template_id": "tmpl_abc123"
}
 
# Pause a scan
POST /api/v1/scans/:scan_id/pause
 
# Resume a scan
POST /api/v1/scans/:scan_id/resume
 
# Stop a scan (preserves results)
POST /api/v1/scans/:scan_id/stop
 
# Cancel a scan (discards results)
POST /api/v1/scans/:scan_id/cancel
 
# Get scan history
GET /api/v1/scans/history

Scan Scheduling

Schedule recurring scans on any connector:

1. Navigate to the connector detail page.
2. Open the Schedule tab.
3. Configure:

Resume Behavior

If a scheduled scan doesn’t complete within its time window:

• The scan checkpoints at the last completed resource
• On the next scheduled run, it resumes from the checkpoint (not from scratch)
• If the checkpoint is older than the configured gap limit (default: 48 hours), the scan restarts fresh to avoid stale coverage

Scan Results

Every scan (running or terminal) includes a structured coverage report so you can audit exactly how much of the target data was covered, what was skipped, and why:

Partial Scan Outcomes

A scan transitions to Partial status (not Failed) when it produced useful findings but did not meet its coverage target, or some resources errored mid-scan. The partial_reason field on the scan record explains exactly why:

Severity Scoring

Every finding is assigned a severity level based on the combination of PII type and detection confidence:

Severity levels are visible throughout the dashboard — in the Data Catalog, Investigation view, Command Center cards, and scan history. Use severity filters to prioritize remediation of the most sensitive exposures first.

Finding Lifecycle

Findings progress through a defined lifecycle as your team triages and remediates them:

To change a finding’s state, select one or more findings in the Data Catalog or Investigation view and use the Actions menu. Bulk state changes are supported for efficient triage workflows.

Per-Object Findings

Every scan result includes a per-object breakdown showing exactly which files, tables, or resources contain sensitive data:

Click any object to drill down into its individual findings with:

• PII type filter — Show only SSNs, only emails, etc.
• Confidence scores — How certain the detection is
• Detection method — Column sampling, full scan, or API pagination
• Masked evidence — Redacted preview (e.g., ****-**-6789) confirming the match

Finding Feedback

On any finding in the Data Catalog, you can provide feedback:

• Confirm (thumbs up) — This is real PII. Strengthens classifier confidence.
• Reject (thumbs down) — This is a false positive. Helps improve detection accuracy over time.

Feedback is used to improve classifier calibration. The more feedback provided, the more accurate future scans become.

Learn More

• Parallel Scanning Engine — Distributed scan architecture and worker scaling
• Classifiers — Managing classifier sets and suppression rules
• Event-Driven Scanning — Real-time scan triggers from cloud storage events
• Run Your First Scan — Step-by-step guide for your first scan