The Problem

Modern website operations require 24/7 monitoring, quick incident response, and careful remediation. Small teams and solo developers struggle to:

  • Detect anomalies before users report them
  • Diagnose root causes from complex logs
  • Execute fixes safely without breaking things
  • Maintain uptime without burnout

The Solution

SitePilot is an AI-powered operations agent that automates the entire incident lifecycle:

  1. Monitor: Continuously tracks HTTP status, response time, SSL certificates, and content changes
  2. Detect: Uses Qwen Cloud AI to identify anomalies (downtime, performance degradation, SSL issues, content tampering)
  3. Diagnose: Qwen analyzes server logs to pinpoint root causes and suggest specific fix commands
  4. Remediate: Automatically executes safe fixes (clear cache, reload nginx) and queues high-risk actions for human approval
  5. Notify: Multi-channel alerts (email, Slack, DingTalk)

How We Use Qwen Cloud

Qwen Cloud is the brain of SitePilot:

  • Anomaly Detection: Qwen-plus analyzes monitoring metrics against historical baseline to detect subtle anomalies that rule-based systems miss
  • Log Analysis: Qwen-plus parses server logs, identifies error patterns, pinpoints root causes, and generates specific fix suggestions (e.g., "run systemctl status postgresql")
  • Natural Language: All AI outputs are in natural language, making operations accessible to non-experts

Key Features

  • Hybrid Detection: Rule-based fast detection + AI deep analysis
  • Human-in-the-Loop: High-risk operations (restart service, block IP) require human approval
  • Fallback Model: Automatically falls back from qwen-plus to qwen-turbo on errors
  • OpenAI-Compatible: Uses Qwen Cloud OpenAI-compatible API for easy integration

Tech Stack

  • Backend: Python 3.11, FastAPI
  • AI: Qwen Cloud API (qwen-plus, qwen-turbo)
  • Storage: SQLite
  • Monitoring: httpx (async HTTP client)
  • Scheduling: APScheduler
  • Notifications: aiosmtplib (email), Slack/DingTalk webhooks

Demo

Built With

  • ai-operations
  • anomaly-detection
  • fastapi
  • human-in-the-loop
  • openai-compatible-api
  • python
  • qwen-cloud
  • sqlite
Share this project:

Updates