AI that handles the volume.
You make the calls.
Smart maintenance triage, AI-drafted tenant replies, and real-time urgency detection — working inside your own permissions, with the high-stakes actions locked to a human.
From request to resolution in minutes
Tenant submits a request
Via the portal, email, or SMS. The message lands in a unified thread.
AI analyzes instantly
Category, priority, urgency, and a draft reply are generated in seconds.
Your team reviews
Approve with one click, edit the draft, or write your own. Tenant messages wait for you unless you've chosen otherwise.
Tenant is updated
Response delivered via their preferred channel. Thread stays unified.
Everything the AI handles for you
Smart Ticket Triage
Every incoming maintenance request is instantly classified by category, priority, and urgency — no manual sorting required.
- Auto-assigns EMERGENCY, HIGH, MEDIUM, or LOW priority
- Detects safety hazards like gas leaks and flooding
- Suggests the right vendor category to assign
- Confidence scores so your team knows when to double-check
Smart Ticket Triage
AI-Drafted Replies
The assistant reads the whole conversation and drafts a reply that sounds like your team — not a bot. By default the draft waits for a human before it reaches a tenant.
- Understands full conversation history
- Adapts tone and length per channel (SMS vs email)
- Built-in guardrails — never promises dates or admits liability
- One-click approve, or reject and write your own
AI-Drafted Replies
Urgency Detection
Every inbound message is scanned in the background for safety risks and time-sensitive issues. Critical situations surface immediately.
- Identifies flood, fire, gas, and electrical hazards
- Detects security concerns like broken locks
- Surfaces legal or compliance language for review
- Runs silently — never delays message delivery
Urgency Detection
Bounded by Your Permissions
The assistant acts as the signed-in user and can never do more than that person's role allows. It has no wider access of its own, and every action it takes is stamped as AI.
- Never exceeds the permissions of the person it's working for
- Every autonomous action recorded as AI, with before/after detail
- Retried jobs are no-ops — it can't double-send or double-charge
- It re-checks that an action actually took effect before reporting success
Bounded by Your Permissions
You Set How Far It Goes
Tenant messages are drafted and held for approval by default. From there you decide, action by action, what the assistant may do on its own — and a set of high-stakes actions stays locked to a human no matter what you choose.
- Widen or narrow autonomy per action, or per property
- Grant time-limited authority that expires on its own
- Opt in to end-to-end auto-send for routine 1:1 tenant messages
- Fair-housing flags, screening, adverse-action notices, binding lease sends, mass sends and vendor spend over your cap can never be automated
You Set How Far It Goes
Three ways to stop it
Triage is applied as it arrives so nothing waits in a queue to be sorted — and stays editable. For anything the assistant does, you always have a way to stop it, at whatever scope you need.
Master switch
Turn the assistant off for the whole organization. Every autonomous behaviour stops — drafting, reminders, follow-ups, the phone line.
Pause autonomy
Keep the assistant available for drafting and answers, but stop it acting on its own. Everything routes back to a human.
Pause one item
Stop autonomous handling on a single ticket, lease or task without changing anything else.