AI Monitoring

AI Monitoring analyzes Drupal logs with AI, sends alerts across channels, and shows a health dashboard with regular status reports—so site owners can spot issues fast and reduce noise.

AI Monitoring helps site owners and teams detect important problems quickly without manually reviewing logs. It captures Drupal log events, analyzes them with AI, applies alert policies, and delivers actionable notifications with clear next steps.

Features

• Captures Drupal log entries and analyzes them in batches with AI.
• Fallback analysis when AI provider calls fail (resilience-first behavior).
• Multi-channel alerting: Email, Slack, Webhook, Drupal notifications.
• Severity-based routing rules for who gets notified and when.
• Noise controls: deduplication, rate limiting, business hours, maintenance window, digest mode.
• Escalation workflow for unacknowledged urgent alerts.
• Dashboard with operational metrics, suppression reasons, and AI analysis state.
• Interval AI reports (6h / 12h / 24h) with trend-oriented summaries.
• Analysis and alert history logs for auditing and troubleshooting.

Use cases: production incident awareness, owner-friendly health updates, reducing alert fatigue, and getting actionable recommendations from technical logs.

Post-Installation

1. Enable the module.
2. Configure an AI provider in Drupal AI.
3. Go to /admin/config/ai/ai-monitoring and enable AI Monitoring.
4. Set severity threshold, batch size, routing, and alert channels.
5. Test your channels (Email/Slack/Webhook/Notification).
6. Review the dashboard at /admin/reports/ai-monitoring.

Cron should be enabled so queue processing, cleanup, escalation checks, and interval reports run automatically.

Additional Requirements

• Drupal core 10.2+ or 11.x
• Drupal AI
• Key
• An AI provider module compatible with Drupal AI (for example OpenAI provider)
• Configured cron (recommended for production)

Recommended modules/libraries

• AI provider modules (for your preferred LLM vendor)
• Syslog / external log shipping modules (if your team also wants centralized observability)
• Incident/on-call platforms via Webhook channel (PagerDuty, Opsgenie, etc.)

Similar projects

Many modules expose logs, watchdog views, or channel notifications, but AI Monitoring focuses on AI-assisted triage + actionable summaries + policy-driven suppression/escalation in one workflow. It is designed for site owners who need fast signal, not raw log noise.

Community Documentation

TBD