Wachd

Wachd transforms on-call response by delivering root cause analysis alongside alerts instead of just firing notifications. When an alert triggers at 3am, your engineer receives not just "HighErrorRate firing" but a plain-English diagnosis correlating recent commits, error logs, and metric history—powered by your choice of AI backend.

Key capabilities include:

• AI root cause analysis - Correlates commits, logs, and metrics into actionable insights using Ollama, Claude, OpenAI, or Gemini
• Flexible on-call scheduling - Primary and secondary rotations with time-window restrictions and self-service overrides
• Per-user notification rules - Email, SMS, voice calls, and Slack in your preferred order with intelligent escalation
• Air-gapped ready - Run Ollama in-cluster with zero outbound API calls for isolated and classified environments
• PII sanitization - Automatically strips emails, IPs, API keys, and account IDs before AI processing
• Multi-tenant isolation - Complete data separation between teams with enforced query-level isolation
• Read-only by design - Connects to infrastructure for analysis only; cannot deploy, restart, or modify anything
• Self-hosted on Kubernetes - Helm chart deployment with support for external or in-cluster databases