Question → Understand → Answer Or Open Request → Route → Confirm

Resident Concierge

Answers resident questions from approved sources, opens requests when needed, and routes anything it should not handle to the right team, so routine questions stop interrupting staff all day.

The Problem

Every question is a phone call. Staff never get a clear hour.

High call volume. Same questions asked repeatedly. Staff interrupted constantly.

What Changes

Answered, opened, or routed, without interrupting the team.

Answers resident questions from approved sources, opens requests, and routes issues to the right team. Your staff handles exceptions, not repetition.

Humans stay in control: Your team handles anything the agent routes, and approves the sources it answers from. It removes the routine interruptions, not the service.

HOW THE AGENT WORKS

TRIGGERLLMAPIDECISIONACTIONWAIT
TRIGGERResident ContactsQuestion or request, any channel
LLMUnderstand IntentQuestion, request, or complaint
DECISIONCan Answer?
LLMAnswer From SourcesReply using approved knowledge
APIOpen RequestCreate a ticket when action is needed
LLMRoute To TeamSend anything sensitive to a person
ACTIONConfirm To ResidentAcknowledge and set expectations
The agent doesn't replace your team. It answers what it can from approved sources and routes the rest. Anything it's unsure of goes to a person.

HOW TO BUILD IT

Three levels of depth. Pick the one that matches where you are.

For the ops person who needs to understand what's involved and brief their team.

Your ResidentsCalls, texts, portal
The AgentWorkflow engine + LLM
Your TeamTickets, routed cases

WHAT YOUR TEAM PROVIDES ONCE

  • Approved knowledge - the policies and FAQs the agent answers from
  • Request rules - when a question becomes a ticket
  • Routing map - what goes to which team or person
  • Escalation triggers - topics the agent must hand to a human
  • Tone and channels - how to respond and where
What happens day-to-day: Residents ask the same routine questions all day. The agent answers the ones it can from your approved sources, opens a ticket when something needs doing, and routes anything sensitive to a person. Your team gets fewer interruptions and cleaner handoffs.

For the IT person or technical lead who needs to scope the integration.

INQUIRY SOURCES

Phone / SMS
Email
Resident portal

WORKFLOW ENGINE (N8N)

Trigger
LLM - Understand
Branch - can answer?
LLM - Answer from sources
PMS API - Open request
LLM - Route
Confirm to resident

CONNECTS TO

PMS (Yardi / AppFolio)Read: resident, lease. Write: requests
Knowledge baseRead: approved policies and FAQs
SMS / email / portalTwo-way resident messaging
Notifications (Slack / email)Route sensitive cases to staff
WITHOUT PMS API

Agent answers from a static FAQ doc and emails the team for anything needing a ticket. Routing logic still runs.

WITH PMS API

Agent identifies residents, answers from your knowledge base, and opens real tickets. Best experience.

For the person who wants to build this agent step by step. Each node below maps to an n8n workflow node.

TRIGGERNode 1
Trigger - Resident Contacts
Twilio voice/SMS + email + portal webhook, normalised.
Output: { resident_contact, channel, message }
LLMNode 2
LLM - Understand & Decide
Classify intent and decide answer vs open-request vs route.
Output: { intent, action, topic, sensitive }
You triage a resident contact.

Return:
- intent: question / request / complaint
- topic: short label
- action: answer / open_request / route
- sensitive: true if it involves money disputes, legal, safety, or harassment (always route)

Respond ONLY with valid JSON.

MESSAGE: {message}
INPUT
"What day is trash pickup on my street?"
OUTPUT
{
  "intent": "question",
  "topic": "trash schedule",
  "action": "answer",
  "sensitive": false
}
DECISIONNode 3
Condition - Can Answer?
IF action=answer and not sensitive → answer from sources. IF request → open ticket. IF sensitive/route → hand to team.
Output: Branch: ANSWER / OPEN REQUEST / ROUTE
LLMNode 4
LLM - Answer From Approved Sources
Answer only from the approved knowledge base; if not covered, route instead of guessing.
Output: { answer, source_ref, answered }
You answer a resident question using ONLY the approved knowledge base.

If the answer is in the KB, return answered=true with the answer and source_ref.
If it is NOT clearly in the KB, return answered=false (do not guess) so it gets routed.

Respond ONLY with valid JSON.

KB: {knowledge_base}
QUESTION: {message}
OUTPUT
{
  "answer": "Trash pickup for your building is Tuesday and Friday mornings. Bins out by 7am.",
  "source_ref": "kb/trash-schedule",
  "answered": true
}
APINode 5
HTTP Request - Open Request
When action needed, create a ticket (optionally via the intake agent) with resident, unit, and detail.
Output: Ticket created for the resident’s request.
ACTIONNode 6
Route & Confirm
Route sensitive or unanswerable cases to the right team, and confirm to the resident what happens next.
Output: Resident acknowledged; staff looped in where needed.
MVP vs Advanced — Where To Start
MVP — START HERE
  • Single channel (SMS)
  • Intent classification
  • Static FAQ answers
  • Email to team for requests
  • Manual routing
  • Fixed ack message
ADVANCED — ADD LATER

    Full Video Walkthrough Coming Soon

    From a resident question to the right outcome - step by step.

    Register to be notified

    Everything on this page is free to learn.

    When you're ready to build — do it yourself or bring us in.

    OTHER AGENTS IN LEASING & MARKETING

    Translates two-way messages, keeps consistent records
    Answers policy questions from approved sources only
    Sends milestone updates automatically, escalates stalled cases
    Sends status updates tied to ticket milestones, automatically

    Frequently asked questions

    Resident Concierge is an AI agent that answers questions, opens requests, and routes to the right team.
    Every question is a phone call. Staff never get a clear hour. The Resident Concierge addresses this — it answers questions, opens requests, and routes to the right team.
    No — your team stays in control. Your team handles anything the agent routes, and approves the sources it answers from. It removes the routine interruptions, not the service.