Skip to main content

Alert Definition

Alerts are a core part of the Coolr platform and support management-by-exception. The alert framework enables administrators to define rule-based triggers, routing, and delivery channels so relevant teams receive timely, actionable notifications.

This page provides a concise overview and links to the detailed category documents:

  • System Alerts: device and data health, telemetry, and pipeline issues — see System Alerts.
  • Merchandising Alerts: shelf, SKU and planogram issues for merchandising operations — see Merchandising Alerts.

Key Concepts​

  • Triggers: conditions based on telemetry, image analysis, or business metrics (e.g., no ping, high rejection rate, out-of-stock).
  • Filters: dimensions used to scope alerts (market, channel, classification, asset type).
  • Recipients: roles or individuals who receive notifications via configured channels.
  • Lifecycle: alerts are inserted, updated while conditions persist, and closed when resolved.

Alert Categories (summary)​

  1. System Alerts

    • Focus: device connectivity, data quality, image pipeline, and operational health.
    • Typical triggers: missed communication, high rejection rates.
    • Details and examples: see System Alerts.

  2. Merchandising Alerts

    • Focus: shelf-level and SKU-level merchandising compliance and stock issues.
    • Typical triggers: out-of-stock by shelf/category, foreign products, planogram deviations.
    • Details and examples: see Merchandising Alerts.

Filters & Notification Routing​

Alerts can be configured using combinations of:

  • Market, Channel, Classification, Asset Type
  • Alert Category and specific Alert Type
  • Granularity: global, market-specific, or location-specific

Routing options:

  • Channel-based: Email, Slack, webhooks
  • Role-based: named roles (e.g., Regional Manager) or individual users
  • Delivery mode: aggregated summary (periodic) or per-event notifications

Example: Configure Grocery channel in North-East to notify regional ops on stock < 80%, and notify Virginia store reps on stock < 85%.

Recipient & Delivery Configuration​

Recipients and delivery rules support:

  • Category-level defaults with per-market overrides
  • Role and individual overrides
  • Multiple channels per recipient (email + WhatsApp)
  • Summary vs. individual event modes (daily digest vs. immediate alert)

Use role-based groups for operational consistency; prefer role channels for high-frequency alerts to avoid individual overload.

API Integration Guidance​

  • Full programmatic configuration is possible but complex; handle with care.
  • Recommended approach: use API only after validating a configuration in the UI or automated tests.
  • See the System Alerts document for available API schemas, payload examples, and required fields.

Best Practices​

  • Start with conservative thresholds and tune using historical telemetry.
  • Use role-based subscriptions for noisy alerts; escalate to individuals for critical issues.
  • Aggregate non-urgent alerts into daily summaries.
  • Maintain a mapping of alert IDs to remediation runbooks for faster resolution.
  • Add tests for any programmatic alert configuration to prevent misrouting.