Help

What does the agent do without me?

It runs a daily sweep over the portfolio: posting each period's rent to the lease ledger, answering residents about their own balance and lease, classifying new maintenance requests and matching them to a vendor from your list, sending the rent reminders on the collection ladder, and drafting the things you will edit — renewal terms, listing copy, the owner summary.

It also raises flags rather than filling gaps in: no late-fee policy configured, no renters insurance on file, a work order that has sat scheduled for too long.

What needs my approval?

Anything that moves money, changes a lease, or penalises a resident. In practice that is late fees, returned-payment (NSF) fees, payment plans, lease renewals, move-outs, security-deposit returns, applicant approvals and denials, vendor dispatch above $400, recurring ancillary fees above $400, posting a utility rebill, and changes to your organisation's own policy settings.

Each one arrives as a task with the amount and the reasoning attached. Approving it is the step that applies the effect — nothing is applied first and unwound later. Declining records the decline.

Can the agent move money on its own?

No. The strongest thing the collection logic can do about unpaid rent is propose a late fee, which becomes a task waiting on you.

Separately, no card or bank rail is connected in the current build at all: recording a payment writes to the ledger, it does not move real funds. Subscription billing and owner payouts are not connected either. That is listed in full on the disclosures page.

How does rent follow-up work — will it hound my residents?

It is a fixed five-rung ladder anchored to the due date, not a loop that gets louder: a reminder three days before rent is due, a notice on the due date, a follow-up when grace runs out, a firmer notice three days after that, and only then a late-fee proposal five days past grace.

One rung per run, never a skipped rung, never two automated messages to the same lease inside twenty hours, and nothing at all outside your quiet hours. A payment plan pauses the ladder entirely, and a late fee you declined is never proposed again for that lease.

What can a resident see when they message the agent?

Their own unit and their own lease. The agent's tools are scoped by role and pinned per instance, so a resident asking about a balance gets theirs and cannot reach another resident's record.

Owners are read-only by design: the agent reports to them, it never acts on their behalf.

Do I have to use a large language model?

No. The language-model brain is off by default. Unless it has been switched on for your deployment, the agent's replies come from deterministic rules.

Either way the approval line is the same — the guardrails are not implemented in the prompt.

How do invites work?

Owners, residents, and maintenance staff do not create their own accounts. A manager invites them by email from inside the product, and they follow a single-use activation link that expires after seven days. Claiming the link is what turns the record into a login.

The invite is scoped: a manager can only invite an owner whose portfolio overlaps their own properties, and only a resident on a live lease. A resident who has moved out cannot claim a link — and the link is not burned, so a corrected lease revives it.

Beta invites for prospective managers are a separate flow, issued from the waitlist.

How do I get an account?

You can create one from the Sign up link, but it is approval-gated: verifying your email queues the request and an administrator still has to let you in. Until then you land on a pending-approval screen rather than in the product.

If you would rather talk first, request early access and tell us about your portfolio.

What is not live yet?

Payments, e-signature, SMS, and applicant screening are simulated in the current build, and outbound email is recorded in the message trail rather than delivered unless a real delivery path has been switched on for the deployment.

We would rather say so plainly than let the product imply otherwise. The full list is on the disclosures page.

Still stuck?

If your question is about how the agent behaves rather than about your account, the how it works page walks the whole loop, and the disclosures page lists what is automated and what is still simulated.