AP automation that checks the bill, not just captures it.

It checks every bill and then does the keying. Each invoice is compared against what that vendor billed for that property before, against your approval thresholds, and against every invoice you already hold from them, with nothing to configure. It also extracts the invoice fields with a per-field confidence score, applies your rules to code the invoice to the right property and GL account, routes it for approval, and posts the approved bill to your ledger — QuickBooks Online and Xero through direct connectors, or a CSV whose columns you define yourself for anything else. Documents get in however they already do: a vendor-facing address on your own intake domain, an upload, or forwarding from the mailbox you use today.

No. inletAP is deliberately not a PMS and not a general ledger. It sits in front of the system you already run and hands it finished, coded, approved bills. That is the whole point of the product — the inbound invoice workflow is the part most property tools do badly.

Not directly, and no vendor can. Mail addressed to a domain you own is routed by your own DNS and mail provider. inletAP receives mail on domains it controls, on your organization's *.inlet-ap.com subdomain. To bring in mail from an address you already publish, set up auto-forwarding at your own provider — Google Workspace, Microsoft 365 and the rest all support it — pointing at whichever inletAP mailbox should handle it. Forward a per-property address to that property's mailbox and the routing follows the mailbox. If you would rather sort on who sent the invoice, a rule on source email can set the property and GL account after it arrives.

Signing up creates the workspace and issues your first intake mailbox in the same session. From there you can forward a real vendor invoice and watch it come out the other side. Vendor mappings, coding defaults and approval tiers can be added as you go — nothing is blocked on completing them first. A big portfolio does not have to be typed in either: export your property, unit, vendor and entity lists from whatever you run today, match your columns to ours once, and import the file. The preview writes nothing — it shows every row it would create or change, and the old value beside the new one, before you import.

The document goes to the review queue instead of being guessed at. Confidence is scored per field, not per document, so you see exactly which value is uncertain and the evidence it came from. Any document whose vendor, number, date or total scores below 70% goes to that queue rather than straight to approval — it is a fixed bar, not a dial you have to tune — and a value the model inferred rather than read is capped below it, so a guess cannot be approved automatically.

Per processed invoice document, in plan tiers, with the prices published on the pricing page — never per user. Every plan has unlimited users, so approvers cost nothing. The free plan covers 25 invoices a month with no credit card and includes the QuickBooks Online and Xero connectors. Paid plans are $39, $99 and $249 a month, and extra invoices beyond the allowance are charged at $0.35 an invoice on Starter, falling to $0.18 on Pro. Prices are in US dollars and exclude sales tax, which is added where it applies. Paid plans renew automatically at the price shown — every month, or every 12 months on annual billing — until you cancel. Cancel any time in billing settings: the next renewal stops and you keep the period you have already paid for.

Three, and only three. QuickBooks Online and Xero each have a direct connector: approved invoices post as bills, with the vendor or contact, the GL account and the tax mapped from the coding applied in inletAP, and the source PDF attached to the bill where the invoice arrived with one. Connect both and one approved invoice posts to both, as one bill in each; if one ledger takes it and the other refuses, the document reads partially exported rather than posted. The third is the CSV export, one generic file you map yourself — seven columns by default, up to 64, at document or allocation granularity, with the header text and the field behind each column under your control. It is not a per-product template, so plan on matching your ledger's import format once. Nothing else is built: AppFolio, Buildium, Yardi and Rent Manager have no connector today and no date. Export batches record external IDs so a posted bill can be traced back to the original document, and retries are idempotent so a retry cannot double-post.

Yes. Right after extraction, against every invoice you already hold from that vendor, on nine matching rules with six guards that stop an ordinary monthly bill or a vendor statement being reported as a copy — including the case where the same invoice arrived twice by two different routes. A confirmed duplicate then holds that invoice out of every export until a person decides: it stays approved and leaves in the next run once somebody answers the card, and the rest of the run goes out as normal. That hold is the only thing in the product that blocks anything.

Yes, once there is something to compare with. Each vendor's history is kept per property, and an invoice at least 10% and at least $50 above the usual amount is noted — becoming a finding when a second invoice confirms it, because one high bill is an event and two is a new price. A vendor that has billed an identical amount every month is treated as a contract, where any move over $25 or 1% is worth one line. It needs about five invoices from a vendor for a property before it can compare prices at all, and it says so on the card rather than staying silent. It never holds the bill, never contacts the vendor, and shows no score or rating of any kind.

Both. A split is computed by amount, percentage, even share, unit count or square footage, held in integer cents so the rows add to the invoice total exactly, and saved as a template you can apply to next month's invoice from the same vendor. Where the properties keep separate sets of books, the invoice posts as one bill per QuickBooks company or Xero organization — a bill is scoped to one set of books in both ledgers and neither API has cross-company posting, so the fan-out is the only correct shape. Split coding is included on every plan, including the free one.