Self-Service Portal Agent · Use case
How can we use an AI agent to answer customers from our knowledge base on our own infrastructure?
A self-service portal agent lets customers ask in plain language, answers from your knowledge base with the article it used, and walks them through troubleshooting step by step. When the articles do not cover the problem, or the customer asks, it passes them to a person with the conversation. It runs on your own infrastructure.
The answer is in an article.
The customer cannot find it.
- ✗ 01Customers cannot find the article that answers them, so they raise a ticket.
- ✗ 02A wrong answer sends a customer down the wrong path.
Where it stops for a person
When the articles do not cover it, or the customer asks.When the articles do not cover the problem, when a step fails, or when the customer asks for a person. It does not guess.
It answers and cites.
A person takes over when needed.
Step · 01
Your knowledge base is connected and indexed in the Knowledge Engine.
Step · 02
A customer describes the problem.
Step · 03
The agent answers from the articles and cites the one it used.
Step · 04
It guides the customer through the steps and checks that each one worked.
Step · 05
Where the articles run out, it passes the customer to a person with the conversation.
Reads from · connectors
SharePoint
Confluence
Notion
Your own APIs
The agent connects to SharePoint, Confluence or Notion for your knowledge base.
Acts through · MCP servers
It hands a conversation to staff through your service system, which it connects to through your own APIs. It changes nothing on an account.
What you set up
- Connect the knowledge base.
- List the subjects that go straight to a person.
- Test with real customer questions, including ones the articles do not cover.
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.
What happens when no article covers the problem?
The agent says so and passes the customer to a person.
Can a customer ask for a person?
Yes, at any point.
Where are customer questions processed?
On your own infrastructure, with a model you serve. Nothing is sent to a public AI service.