Blog
Tutorials & notes
· 5 min read
Haiku vs Sonnet: which one is better for incident alerts?
Does an AI model find the real cause of a failure, or only sound like it did? We staged failures in the OpenTelemetry Demo and graded Claude Haiku 4.5 and Claude Sonnet 5.5 on 37 investigations.
· 4 min read
Jev and AlertINT learned the same lesson about AI
A new AI model called Jev never writes a sentence. That sounds like a step back. It's the same lesson we learned building AlertINT: AI works best inside clear rules.
· 5 min read
Tutorial: AI triage for Zabbix alerts, on your own infra
Wire Zabbix into AlertINT in about 30 minutes: problems become correlated, LLM-triaged incidents your AI coding agent can read over MCP — read-only, self-hosted, nothing leaves your network.