OT Security Agent · Use case
How can a manufacturer use an AI agent to triage alerts on its operational technology network on its own infrastructure?
An OT security agent takes an alert about an unusual command or unauthorised access on the operational technology network, gathers the logs around it and past incidents like it, and writes up what it found for the security team. The team decides and responds: the agent sends no command to plant equipment. It runs on the manufacturer's own infrastructure.
Plant networks were built to run.
Not to be watched.
- ✗ 01Plant networks carry commands that can stop a line or damage equipment, and few people can read their logs.
- ✗ 02Details of the plant network must not leave the company.
Where it stops for a person
Always, before any response on the plant network.Always. The security team and plant engineers decide the response.
It gathers the context.
The security team responds.
Step · 01
An alert arrives from your monitoring.
Step · 02
The agent reads the logs and metrics around the alert.
Step · 03
It checks whether the command or access fits the normal pattern for that equipment.
Step · 04
It writes up what it found, with the entries it read.
Step · 05
The security team reviews the write-up and responds.
Reads from · connectors
The agent connects to PagerDuty and ServiceNow for alerts and past incidents, and to Grafana or Prometheus through MCP servers for logs and metrics.
Acts through · MCP servers
Grafana
Prometheus
Nothing. It gathers the evidence and reports. It has no access to plant equipment.
What you set up
- Connect the alerts and the monitoring tools.
- Describe what is normal for each type of equipment.
- Test against past incidents.
Where it runs
The agent, its knowledge store and the model run on infrastructure you own or control. You choose the model and serve it yourself, so questions, documents and answers are not sent to a public AI service.
How you get it today
This page is labelled Use case: the agent is built in Katonic Studio. You describe the job, connect what is named above, add test questions and release the agent once it passes them. When the Catalog holds a tested template for this job, this page will carry the Template label and show how to take it.
Can the agent send commands to plant equipment?
No. It has no access to it. It reads logs and reports.
Does it replace our OT monitoring?
No. It works on the alerts your own tools raise.
Where are plant network logs processed?
On your own infrastructure, with a model you serve. Nothing is sent to a public AI service.