Skip to content
← All work
Case study · n8n · LLM · Trivy

An AI agent on the on-call rotation

n8n agent workflows that watch K8s clusters, enrich alerts with logs and scan results, and send SREs a diagnosis with the fix.

Company
BerryBytes
My role
DevOps Engineer
Timeline
2024 — present
Type
AI · DevSecOps
Stack
n8n, LLM, Trivy
01
Alert firesPrometheus / Cortex
02
n8n triggerAlertmanager webhook routes by severity & namespace
03
AI agentPulls logs, events & Trivy scans as tools
04
DiagnosisRoot cause + runbook fix
05
Notify SRESlack / email with context attached
01 · What it is

AI On-call Agent, in one sentence

n8n agent workflows that watch K8s clusters, enrich alerts with logs and scan results, and send SREs a diagnosis with the fix.

02 · What I did

DevOps Engineer at BerryBytes

  • When a Kubernetes alert fires, an n8n agent collects context, reasons over it with an LLM and sends the engineer a diagnosis and suggested fix.
  • The agent pulls logs, events and Trivy scan results as tools before it answers.
  • Workflows route by severity and namespace, so the right engineer gets the right context.
the part I'm proudest of

The engineer gets the diagnosis and a suggested fix before they open a dashboard.

03 · Under the hood

The stack

n8n, LLM, Trivy — ai · devsecops work, built to be reproducible and observable from day one.

Next projectHoryzon Pictures streaming →
Your turn

Want a platform like this?

Tell me what you're running today and where you want to get to.

Start a project