What is this?
Keep is an open-source AIOps and alert management platform: a single pane of glass for alerts and incidents, with deduplication, enrichment, filtering, correlation, workflows and dashboards built in.
Is this a cloud service I sign up for, or something I run myself?
You run it yourself: it's a self-hosted platform with its own customizable web UI for alerts and incidents, and the repo ships a docker-compose file to bring up the stack.
How do I get it running?
The repo includes a docker-compose file, so the typical path is cloning the repo and bringing the stack up with docker compose; the full platform documentation walks through the whole setup.
Which monitoring tools can it pull alerts from?
Dozens of observability tools, synced bi-directionally: Datadog, Grafana, Prometheus, CloudWatch, Sentry, Splunk, Zabbix, New Relic, Dynatrace, Elastic and many more.
Does it replace my Datadog or Grafana, or sit on top of them?
It sits on top of your existing stack: Keep syncs bi-directionally with the monitoring tools you already run, so they keep collecting while Keep deduplicates, correlates and manages the alerts.
What can I automate with workflows?
Think GitHub Actions for your monitoring tools: customizable automated workflows that act on your alerts and incidents instead of manual triage every time something fires.
Can the AI features run locally, without sending alert data to a cloud API?
Yes — besides Anthropic, OpenAI, DeepSeek, Grok and Gemini, Keep supports Ollama and LlamaCPP as AI backends, so enrichment and correlation can run on models hosted on your own hardware.
Can it notify my team in chat?
Yes — it integrates communication platforms including Discord, Google Chat, Mattermost, Pushover, Ntfy.sh, Mailgun and Resend, alongside the observability and database providers.
Can it pull context from my databases?
Yes — supported database and warehouse providers include BigQuery, ClickHouse, Databend, MongoDB, MySQL, PostgreSQL and Snowflake, on top of the monitoring integrations.
Which topics is it in?
monitoring-tool, sre, alerting, incident-response, devops-tools, python3, notifications, workflow-automation, local-llm, observability, ollama, docker-compose