> ## Documentation Index
> Fetch the complete documentation index at: https://zenorent.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Bookkeeper

> Ask questions, review bookkeeping data, and prepare approved actions through chat.

The AI Bookkeeper is a chat assistant that reads your data, prepares bookkeeping actions, and waits for your approval before it changes records.

## Where to use it

* **In the app** — open **Bookkeeper** under **Zeno** in the sidebar. On mobile, the same entry is labeled **AI**. Conversations are saved in the sidebar for later.

* **By phone or WhatsApp** — an optional add-on, set up by us. See below.

## By phone and WhatsApp

The assistant answers actual phone calls. Someone rings the number, asks what a tenant's balance is, and hears the answer — no app, no login. WhatsApp works the same way in text.

It speaks **English and Hebrew**, and picks up which one it's hearing. Hebrew is handled by a speech engine chosen for it specifically, because the general-purpose one transcribes it poorly.

Over the phone it can look up tenants, balances, payment history, invoices, recent payments, and a property summary — the same reading it does in the app. It doesn't post transactions by voice.

Access is by caller ID. We authorize each phone number against a person in your organization, and calls from anywhere else are refused. Voice and WhatsApp are enabled per person, so you can give a property manager phone access without giving it to everyone.

> [!NOTE]
>
> There's no self-serve setup for this yet — ask us to turn it on and tell us which numbers to authorize.

## What else lives in the Zeno section

Underneath Bookkeeper, the sidebar has two companion surfaces:

* **Knowledge & rules** — everything Zeno knows about your books, on three tabs. **Entries** is the org-wide memory the Bookkeeper draws on: routines, categorization rules, and review hints saved from past chats, with confirmed entries auto-injected into future tool results. **Rules** is the bank-feed auto-categorization rule list. **Log** is a session-by-session record of what the Bookkeeper did, what tools it called, and what you approved.
* **Inbox** (Beta) — a queue view of bank-feed items the Bookkeeper has clustered into draft work packs for batch review.

## What it can look up

* Tenant balances, payment history, invoice history
* Properties, units, leases, and rent amounts
* Items in the Inbox and documents that have been filed
* Recent payments across the organization
* QuickBooks customers, invoices, payments, and credit memos (if connected)
* Comparisons between Zeno and QuickBooks

## What it can prepare

For write actions, the Bookkeeper validates your request and shows the details before it acts. You approve the action in chat; the audit trail records it under your user with agent context.

* **Categorize bank-feed items** — post one or more pending items to the ledger with a GL account, vendor, memo, and cost-center scope.
* **Skip bank-feed items** — mark items as ignored with a reason.
* **Create or update rules** — preview a rule's shape and match count, then save it after you approve.
* **Search documents** — find uploaded scans, receipts, bills, statements, and filed documents.
* **Review AP and AR** — list bills, invoices, and aging summaries.
* **Check reconciliation state** — summarize a GL account's last reconciliation and uncleared transactions.
* **Save org knowledge** — record approved routines, categorization rules, and review hints for future sessions.

Approval-gated writes require a clear yes from you before the Bookkeeper calls the tool.

## Tips for fast answers

* Use full tenant names or unit numbers.
* Include the amount for payments and credits.
* Say "yesterday" or a specific date if it's not today.
* Follow up in the same conversation — the Bookkeeper keeps context.

Invoice workflows, lease edits, full Inbox review, and accounting sync still happen in the main app.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.