WORK / HyperMonitors
2025
CASE STUDY · 2025
HyperMonitors
HyperMonitors is a production SaaS for uptime and infrastructure monitoring. It runs seven kinds of check — HTTP with keyword matching, TCP port, ping, DNS, SSL, domain expiry, and inbound heartbeats for cron jobs — from three geographic regions, using cross-region quorum so a single failing probe reads as degraded rather than triggering a false incident. When something does go down, alerts fan out across nine channels including Slack, Discord, Telegram, PagerDuty, SMS, and webhooks, with on-call rotations deciding who actually gets woken. Teams get branded public status pages on their own domains, an audit log, scoped API keys, and an MCP server for querying monitors from AI tooling. Built as a TypeScript monorepo: Next.js frontend on Vercel, Fastify API and check engine in Docker on a VPS, Postgres via Supabase, Stripe for billing, and fully localized in English and German.
OVERVIEW
HyperMonitors is a production SaaS for uptime and infrastructure monitoring. It runs seven kinds of check — HTTP with keyword matching, TCP port, ping, DNS, SSL, domain expiry, and inbound heartbeats for cron jobs — from three geographic regions, using cross-region quorum so a single failing probe reads as degraded rather than triggering a false incident. When something does go down, alerts fan out across nine channels including Slack, Discord, Telegram, PagerDuty, SMS, and webhooks, with on-call rotations deciding who actually gets woken. Teams get branded public status pages on their own domains, an audit log, scoped API keys, and an MCP server for querying monitors from AI tooling. Built as a TypeScript monorepo: Next.js frontend on Vercel, Fastify API and check engine in Docker on a VPS, Postgres via Supabase, Stripe for billing, and fully localized in English and German.
FEATURES
[05]- 01Monitoring - Seven check types: HTTP/HTTPS, TCP port, ping, DNS, SSL, domain expiry, and heartbeat (cron jobs ping in; silence past a grace period alerts) - HTTP keyword matching — catches soft failures that return 200 but serve the wrong content - Custom request headers, so authenticated/private APIs don't read as down - Check intervals down to 30 seconds - Fast-confirm re-check to catch short outages between scheduled checks
- 02Multi-region - Probes in Zurich, New York, and Singapore - Per-region quorum: partial failures show as degraded instead of firing a false incident - Per-region uptime breakdown and filtering
- 03Alerting - Nine channels: in-app, email, Slack, Discord, Telegram, MS Teams, SMS, PagerDuty, webhooks - On-call rotation schedules with escalation to whoever's currently on call - SSL expiry reminders at 30/14/7/3/1 days, plus unexpected-reissue alerts - Maintenance windows (one-off and recurring) that suppress alerts - Weekly uptime digest email
- 04Status pages - Public pages with live status, uptime bars, and incident history - Custom domain via a single CNAME - Branded with your logo and accent color; light, dark, or visitor-following themes - Password protection or team-only visibility - Email subscribers for incident notifications - History windows from 30 days to 2 years
- 05Teams & access - Multi-tenant orgs with role-based access and purchasable extra seats - Audit log of who changed what - Scoped API keys with rate limiting - Full REST API, plus an MCP server for AI-assistant access - English and German throughout, including all emails and alerts
MORE WORK
CONTACT
LET'S TALK
Tell us about your project