Citizen Services Agent · Use case
How can a government agency answer citizens' questions with an AI agent on its own infrastructure?
A citizen services agent answers questions about benefits, permits and public services from the agency's own published guidance, in the languages you choose, and cites the page it used. A question about an individual case goes to a member of staff with the conversation so far. The agent runs on infrastructure the agency controls, so citizens' questions are not sent to a public AI service.
Long queues for simple answers.
In more than one language.
- ✗ 01Citizens wait in a queue for answers that are already on the agency's website, in a language they may not read well.
- ✗ 02A wrong answer about an entitlement misleads a citizen, and citizens' questions cannot leave government infrastructure.
Where it stops for a person
Individual cases and decisions go to staff.Anything about an individual's case, entitlement or appeal goes to a member of staff. The agent does not decide eligibility.
It answers from published rules.
Cases go to staff.
Step · 01
The agency's published guidance is connected and indexed in the Knowledge Engine.
Step · 02
A citizen asks a question in their own language.
Step · 03
The agent answers from the guidance only and cites the page.
Step · 04
Where the question concerns an individual case, it hands the conversation to staff.
Step · 05
A member of staff takes over.
Reads from · connectors
SharePoint
Confluence
Your own APIs
The agent connects to SharePoint or Confluence for published guidance.
Acts through · MCP servers
It hands a conversation to staff through your case system, which it connects to through your own APIs. It changes no record and makes no decision.
What you set up
- Connect the published guidance.
- Choose the languages and the subjects that go to staff.
- Write test questions in each language, including ones it should hand over.
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.
Does the agent decide who is eligible?
No. It explains the published rules. A member of staff decides each case.
Which languages does it answer in?
The ones you choose and test. Each language has its own test questions.
Where are citizens' questions processed?
On infrastructure the agency controls, with a model it serves.