The owner statement, automated: a complete blueprint

Who this is for: anyone managing units for owners — from a broker with five landlords to a PM company with five hundred — who spends the first days of every month turning a spreadsheet into statements.

Three days of month-end Excel is not an accounting problem. It's a formatting problem, and formatting problems are exactly what AI is good at.

1 · The data

Everything starts from one export. Most PMS systems (and most spreadsheets) can produce a monthly CSV with a row per unit. You need at minimum:

• unit — building, unit number, type
• owner — who receives the statement
• status — occupied / vacant, and since when
• rent_collected — actual receipts for the month
• expenses — maintenance and charges, itemised with short descriptions
• notes — anything the team logged (renewals, complaints, works)

Do not clean the data beyond this. The point of the workflow is that it consumes what you already have.

2 · The prompt

Paste the CSV into a capable AI assistant together with this instruction. Adjust the bracketed parts once, then reuse every month:

You are the reporting desk of [COMPANY], a Dubai property
management firm. Below is this month's portfolio export (CSV).

For EACH owner in the data, draft a monthly statement with:
1. A 2–3 sentence summary of the month in plain, factual
   language (occupancy, income vs last month, anything notable).
2. A table: unit · status · rent collected · expenses · net.
3. An "actions taken" list built ONLY from the notes column.
4. A closing line stating net amount payable to the owner.

Rules:
- Use ONLY figures present in the data. Never estimate.
- If a figure looks inconsistent (negative rent, expense with
  no description), do NOT include it — list it under
  "FOR REVIEW" at the top of that statement instead.
- Currency: AED. Tone: professional, warm, no marketing.
- Flag any unit vacant for more than 14 days.

Output each statement separately, ready to paste into email.

3 · The checks — this is the part people skip

The prompt above already contains the most important control: anything inconsistent goes to a "FOR REVIEW" block instead of into the statement. Add two human checks on top:

• Totals check. The sum of all net amounts across statements must equal the net of the source CSV. Thirty seconds in Excel. Do it every month.
• One full read. Read one statement end to end — a different owner each month. You're checking tone and sanity, not arithmetic.

With these two checks, the workflow is a drafting assistant, not an accountant — and that's exactly the right altitude for it.

4 · The three ways it breaks

• Dirty notes columns. If your team logs "spoke to owner, angry" in the notes, it will end up in a statement. Fix the logging culture or add a rule excluding sentiment from output.
• Mid-month ownership changes. The CSV shows one owner; reality had two. This is a data problem no prompt can fix — handle transfers manually.
• Silent format drift. The PMS update renames a column and the numbers land in the wrong place. The totals check catches this — which is why it's non-negotiable.

5 · What this costs and saves

Cost: roughly the price of a coffee in API or subscription fees per month-end. Saving: for a 50–200 unit portfolio, typically two to three working days per month, converted into a fifteen-minute review. The owners also get better statements — consistent, readable, on time — which is worth more than the saved days.