The product
The full operating surface, in one product.
Telemetry, diagnosis, routing, and forensic history — engineered as one cohesive surface for teams that run their own Linux infrastructure. No agents to glue together. No dashboards to assemble.
Telemetry
Continuous host telemetry
Every signal a Linux operator actually reaches for, captured at second-level resolution.
The agent reports the operational primitives that determine whether a host is healthy: CPU pressure, memory behaviour, disk saturation, load, and the runtime of the services that sit on top. No proprietary formats, no opinion-laden abstractions — just a clean picture of what’s happening on the box.
Diagnosis
Diagnosis engine, not a chatbot
When the page fires, you get a written brief on what broke and why — before you’ve finished opening the laptop.
The diagnosis engine reads metrics, system events, recent service restarts, and config changes the moment an alert opens. It writes a short brief in the language an on-call engineer would use: what failed, what likely caused it, and the first three things worth trying. No prompt engineering required.
Routing
Alerts that respect on-call
Threshold-based routing that doesn’t spam, with deduplication built in.
Alert rules are scoped per host, per environment, or fleet-wide, and route to whichever channel the responsible team actually reads. Cooldowns, grouping, and severity gating are first-class — not afterthoughts in a YAML file.
Forensics
An auditable record of every incident
Postmortems shouldn’t begin by reconstructing the timeline from Slack scrollback.
Every incident is captured end-to-end: detection, escalation, diagnosis, mitigation, and resolution. Metric history sits alongside it, so you can show the exact graph that justified the page weeks after the fact.
Agent
A single static binary
One install command. No package manager fights. No daemons you didn’t ask for.
The agent ships as a single Linux binary under fifteen megabytes of resident memory. It works on Debian, Ubuntu, Alpine, RHEL, and anything else that runs systemd or OpenRC. If your VPS has SSH and curl, it’s supported.
Workspaces
Workspaces sized for the team
Shared visibility for the people responsible. No per-seat invoices for the rest of the org.
Invite the engineers who carry the pager. Group hosts by environment, customer, or service owner. Route alerts so the right person sees them and the wrong person doesn’t.
Trust
Security held to operator standards
Nothing leaves your host that doesn’t need to. Nothing crosses the wire in the clear.
We treat your telemetry the way you’d treat your own production data. The agent uses a scoped token, communicates over TLS 1.3, and never touches application data, secrets, or user content.