Wachd

Wachd is a self-hosted alert management and root cause analysis platform designed as a replacement for OpsGenie, with a focus on privacy and in-cluster data residency. It addresses the needs of on-call engineering teams by not only aggregating alerts from sources like Grafana, Datadog, and Prometheus, but also providing automated root cause analysis using AI models that run entirely within the user's own Kubernetes cluster. This ensures that sensitive incident data remains inside the organization’s infrastructure and never leaves the cluster, supporting air-gapped and regulated environments.

The platform integrates with a variety of monitoring and communication tools, including GitHub, GitLab, Loki, Slack, Microsoft Teams, Twilio, and more. Upon receiving an alert via webhook, Wachd validates the event, collects relevant context such as recent commits and error logs, and analyzes the issue using AI models like Ollama, vLLM, Claude, OpenAI, or Gemini. The analysis produces a concise root cause summary and suggested action in plain English. Notification delivery is flexible, supporting SMS, voice calls, email, and Slack, with configurable escalation policies and per-user notification rules. On-call scheduling includes primary, secondary, and custom rotations, as well as time-window restrictions and self-service overrides.

Security and privacy are central to Wachd’s design. Personally identifiable information such as emails, IPs, API keys, and account IDs are stripped from data before any AI processing. The tool enforces strict read-only access to infrastructure, logs, and repositories, and supports multi-tenant isolation to keep teams’ data separate. Daily CVE breach intelligence scans are cross-referenced with actual dependencies, and the system can auto-open fix pull requests. SSO integration with Entra ID, Okta, and Google Workspace is available, alongside local user accounts for break-glass access.

Wachd is delivered as a self-hosted solution on Kubernetes, deployable via Helm, and supports both in-cluster and external databases. 0 license, with a free tier for unlimited teams and users. Paid plans offer additional features such as cloud AI integrations, incident history exports, and enterprise compliance options. The platform is suitable for organizations requiring robust, private, and customizable alerting and incident response workflows.