加载中
正在获取最新内容，请稍候...
Keep - The Open-Source AIOps and Alert Management Platform
Keep is an open-source platform designed to streamline AIOps and alert management. It helps organizations consolidate alerts from various monitoring tools, reduce noise through intelligent correlation, provide actionable insights using AI, and improve incident response workflows.
Python
Added on 2025年7月6日10,453
Stars1,021
ForksPython
LanguageProject Introduction
Summary
Keep is a modern open-source AIOps and alert management platform built to centralize, correlate, and automate the handling of alerts from diverse sources, enabling faster and more efficient incident response.
Problem Solved
Organizations often face alert storms from fragmented monitoring systems, leading to delayed incident response and wasted engineer time. Manually correlating events across different tools is inefficient and prone to errors. Keep addresses these challenges by providing a centralized, intelligent platform for alert processing and incident management.
Core Features
Unified Alert Ingestion
Ingest alerts from a wide array of monitoring, logging, and security tools through native integrations and webhooks.
Intelligent Alert Correlation
Automatically group related alerts based on various criteria to reduce alert fatigue and identify root causes faster.
Automated Remediation Workflows
Define automated workflows and runbooks to trigger actions, notifications, or ticket creation based on incoming alerts.
Tech Stack
Python
FastAPI
Kafka
PostgreSQL
React
Kubernetes
使用场景
Keep can be leveraged in various scenarios where effective alert management and incident response are critical:
Consolidating Monitoring Tools
Details
Integrate alerts from multiple monitoring systems (Prometheus, Datadog, Grafana, etc.) and logs into a single platform.
User Value
Provides a single pane of glass for all alerts, reducing context switching and improving oversight.
Reducing Alert Fatigue
Details
Automatically group related alerts originating from the same service or incident, filtering out duplicates and less important notifications.
User Value
Engineers receive fewer, more relevant notifications, allowing them to focus on critical issues without being overwhelmed.
Automating Incident Workflows
Details
Define rules to automatically escalate alerts, trigger notifications in collaboration tools (Slack, Teams), or create tickets in issue trackers (Jira, ServiceNow).
User Value
Speeds up the time-to-detection and time-to-resolution by automating initial response steps.
Recommended Projects
You might be interested in these projects
clap-rsclap
This project is an efficient tool designed to automate specific tasks, significantly improving workflow efficiency and accuracy. It's suitable for developers and analysts dealing with large datasets or repetitive processes.
Rust
152311112
tensorzerotensorzero
TensorZero is an open-source stack for industrial-grade LLM applications. It unifies an LLM gateway, observability, optimization, evaluation, and experimentation.
Rust
8102496
raspberrypilinux
Official kernel source tree for Raspberry Pi-provided kernel builds. This repository is primarily for issues related to the Linux kernel itself. For other issues, please use the community forum at https://forums.raspberrypi.com/
C
118125192