An AI agent for municipalities and public services
that answers resident questions, not policy decisions
A municipal office fields the same resident questions daily: what are today's hours, what does a permit application need, where does a service request stand. We build an agent that answers from your actual published rules and schedules, and checks request status on its own. Anything that needs a real decision goes to staff, never answered as if it were policy.
Questions that are already public record
A municipal front desk and phone line field a steady stream of questions that are entirely a matter of public record. What time does the office open. What documents does a building permit need. Where does my service request currently stand. None of these need a staff member’s judgment. Each one still takes a person away from a desk to look up and recite information that is, in principle, already published somewhere.
Availability makes it worse. Residents try to reach a municipal office outside the hours it is staffed, evenings and weekends, and get nothing until the next business day. That holds even for a question as simple as confirming hours or a form’s requirements.
Consistency suffers too. Without one authoritative source a staff member checks, the answer to “what does this permit require” can vary depending on who answers the phone. An inconsistent answer from a public office carries more weight, and more consequence if wrong, than it would from a private business.
There is also a seasonal volume problem tied to known events: a filing deadline, a public notice period, a seasonal service change. The same question then arrives from dozens of residents in a short window, and the office’s staffing does not flex with that pattern.
The agent we build for municipalities and public services
The agent answers office hours, location and permit-requirement questions from your actual published rules and schedules. It checks service request status by reference number, and routes a new request to the right department based on what the resident actually needs. The knowledge base comes from your real published documents, not a generic template, and staff can update it as rules change without any code involved.
The agent has one hard rule in its playbook: it never interprets a regulation, grants an exception, or gives anything that resembles legal advice. The moment a question moves from “what are the requirements” to “can I get an exception,” it says so. It hands off to staff with the full conversation attached, so nothing needs repeating.
Typical integrations: your municipal website’s chat and WhatsApp for resident questions, and a case-management or permit system where an API exists. Smaller offices not yet running a dedicated system can use a structured spreadsheet instead. For jurisdictions handling several service types, permits, utilities, public records, the knowledge base is scoped by department. An update to one department’s rules never risks introducing stale information into another’s answers.
Three to five weeks, step by step
- Audit of current resident inquiries. We read a sample of real questions from phone logs, email or your current contact channels, and map what repeats.
- Playbook. FAQ content pulled from your published rules and schedules, department routing, and the exact line the agent will not cross, reviewed with your office before any code.
- Build and test. The agent is built against your real rules and systems, tested on a sample of real questions first.
- Launch with guards. A clear escalation rule for anything beyond published information, full logging of every handoff.
- Tuning. Two weeks of reviewing real conversations and adjusting the knowledge base.
What it costs
| Package | Price | What it covers | Timeline |
|---|---|---|---|
| Assistant | from $2,000 | One to two channels, FAQ, permit lookups, status checks, routing | 3 to 5 weeks |
| Resident agent | from $4,000 | Multi-department intake, structured request routing, system writes | 4 to 6 weeks |
| Agent team | from $10,000 | Resident agent plus a monitor across several departments or offices | 6 to 10 weeks |
Running cost is separate and usually $20 to $200 a month in model usage depending on inquiry volume, with a hard budget cap set before launch.
What usually moves after launch
Public bodies that automate routine resident FAQ typically see front-desk and phone inquiry volume on repeat questions fall by a meaningful share. Comparable public-sector deployments often report 30 to 50 percent. The agent answers from a fixed, published knowledge base, not from whoever happens to be at a desk.
These are typical ranges reported across the sector, not a promise for every office. Service complexity and resident volume both move the number. Our own numbers sit in the case studies below. A visa consulting centre’s client-facing AI consultant is detailed in the visa centre AI support case study. It uses a strict hand-off rule and an admin-editable knowledge base. An outreach and inbox agent built on a structured playbook for a B2B certification body is in the B2B certification AI sales agent case study.
Why work with us
We build these agents the way we build them for ourselves. A closed set of tools. A hard rule the agent cannot cross, here, no regulatory interpretation or exceptions, enforced in the playbook rather than hoped for. The price is fixed once we agree the plan. You get a working demo every week during the build, and the conversation logs and source code stay in your office’s name.
If your real bottleneck is that rules and requirements are published inconsistently across different documents, we will say so. An agent reading conflicting sources just automates the inconsistency faster. Consolidating that source of truth first is usually worth doing before any agent goes live on top of it.
Pair this agent with a WhatsApp bot for municipalities and public services for lighter appointment and request-status needs. See the full package breakdown on the AI agents service page, or get a written plan with a fixed price for your office.
FAQ
How much does an AI agent for a municipality cost?
Our Assistant package starts at $2,000. That covers one or two channels, FAQ, permit lookups, service request status and routing, live in 3 to 5 weeks. A lighter FAQ-only bot starts at $900.
Does the agent interpret regulations or give legal advice?
No, and that boundary is written into the playbook before launch. It answers from your published rules, requirements and schedules, checks request status, and routes new requests to the right department. Any question that requires interpreting a regulation or making an exception is routed to staff.
How does it know our permit requirements and schedules?
From your published rules, forms and office hours, which we build into the knowledge base. Staff can update it as rules change, without touching code, so the agent never answers from outdated information by accident.
Which channels does it run on?
Your municipal website's chat and WhatsApp are the most common for resident questions, with Telegram as an option for jurisdictions where that channel is widely used.
Can it serve residents in more than one language?
Yes. Jurisdictions with multilingual resident populations typically run the agent in the official language plus one or two commonly spoken languages locally.
Does it integrate with our permit or request management system?
We integrate with common municipal case-management or permit systems where an API or reliable export exists, or with a structured spreadsheet for smaller offices. We tell you upfront if your current system has no reliable way to connect.