A property management business runs on its trust accounting software, a shared inbox and the phones. Rent comes in with whatever reference the tenant typed, maintenance requests arrive as texts with a photo of a dripping tap, and routine inspection reports are written up from notes once the last inspection is done. Property managers and the trust accountant carry most of it by hand, alongside the sales team’s enquiries and client checks. The trust reconciliation, lease and option dates, and complaints land on the same people. Receipting, reminders, routing and register checks are predictable, and those steps run as fixed, tested software. AI is used only where something has to be read or drafted, such as a tenant’s message or an inspection note, and the property manager checks its output before anything goes to an owner or tenant.
These are patterns we see in real estate and property management businesses, not client case studies. Your process gets its own map in the assessment.
Maintenance requests, arrears and inspection reports
A tenant texts a photo of water under the kitchen sink late on a Friday. By Monday the property manager has to judge the urgency, pick the owner’s preferred tradesperson, check the approval limit and arrange access. Arrears run on the same person’s memory, and routine inspection reports get written from phone notes. AI reads the maintenance request and classifies the job and its urgency, and the work order is raised with the owner’s preferred trade inside the approval limit. Anything that looks urgent goes straight to the property manager. Arrears notices follow the cadence you set. For inspections, AI drafts the report text from notes and photos, and the property manager edits it before the owner sees it.
Receipting rent and chasing sundry invoices
Most tenants pay with the right reference. The rest pay under a nickname, send a partial amount, pay from a housemate’s account, or make a single transfer covering rent and a water bill. Those deposits sit unreceipted until someone works them out, and meanwhile the tenant shows in arrears. Commercial tenants add outgoings invoices and recoveries that need chasing. Your trust accounting software already sends basic reminders, so the gain is in the matching. Clear payments are receipted to the right ledger on rules. AI reads commercial remittance advice and suggests where an unclear deposit belongs, and the trust accountant confirms each suggestion before it is receipted to trust. Disputes over outgoings go to the property manager.
Trust money and client checks
Trust money belongs to owners, tenants and buyers, and every reconciliation has to bring the bank, cash book and ledgers into agreement. The trust accountant spends days on unpresented items, unidentified deposits, and the ledger that is out by a small amount nobody can find. The work is high-consequence, so it runs as fixed software only, with no AI near the balances. Bank transactions are matched to receipts and disbursements on rules, each ledger is checked against the bank, and every unmatched item is listed with its history. The licensee in charge reviews the reconciliation and signs it off. The automation keeps the records and runs the routine steps. The trust account remains the business’s responsibility.
AML/CTF obligations started on 1 July 2026 for real estate agents providing designated services. For a sales team that means identity checks on the people it acts for, beneficial-owner checks when the client is a company or trust, a risk rating, and records that can be produced later. For those entities, AI reads the trust deed and company extract to find the people behind them, and staff check what it found. The routine steps run the same way for every new client, calling your identity verification service and filing the results against the listing. The risk rating is a decision for the person you have made responsible, and the agency keeps responsibility for meeting its obligations.
Running the rent roll
The pm@ address gets everything. Tenant repair requests land beside owners asking about rent, tradespeople sending quotes and invoices, rental applications, bond paperwork and the occasional angry email. Property managers sort it between inspections, and messages wait until someone scrolls far enough down. AI reads each message, works out which property and which person it concerns, and classifies it. Routing rules then send invoices to accounts, quotes to the owner-approval step, applications to leasing and repairs to maintenance. Anything the AI is unsure about waits for a property manager to place it, and nothing is guessed. Each property manager opens the day with their own work, already sorted by property.
Commercial leases carry option dates, rent reviews, make-good clauses and notice periods. Residential tenancies have their own end dates and renewal steps, and management agreements with owners have terms of their own. Most agencies track these in the trust software’s diary and a spreadsheet, and a missed option date is expensive for the owner and awkward for you. AI reads each lease and agreement to extract the dates and obligations into a register, and a staff member checks every extracted date against the document before it goes live. From then on the property manager is reminded ahead of each date, early enough to act. What to recommend to the owner stays with the property manager.
Most property management software already sends entry notices and inspection reminders, and does it well. We would not rebuild that. The part still done by hand is everything after the reminder: the tenant who replies that they are away, the owner who wants to attend, the tradesperson who needs a key, and the rebooking that ripples through the property manager’s inspection run. Automation picks up those replies. AI reads the tenant’s response and proposes a new time within the notice rules you have configured, and fixed software rebooks the run and issues the new notice. Changes that affect an owner or a trade are confirmed by the property manager first.
Enquiries and complaints
A complaint in property management can arrive as an email to the principal, an online review, a call to reception or a note on a maintenance job. Often nobody recognises it as a complaint until it has escalated. Every complaint goes into a single register, linked to the property and the ledger, with a response date and an escalation when that date passes. AI drafts a response from the file history, and the property manager or principal approves it before it is sent. The principal can see which properties, trades and staff keep appearing, and nobody has to reconstruct a file when a complaint reaches a tribunal.
Enquiries come from the property portals, the website appraisal form, open-home sign-ins and phone calls, and many never make it into the CRM properly. Your CRM and the portals already connect for much of this, so the gap is usually narrower than it looks. Where it isn’t covered, fixed software brings each enquiry into the CRM, matches it to an existing contact or creates a new one, attaches the property and assigns it to the right agent on your rules. Open-home sign-in sheets and forwarded emails get the same treatment. Agents keep the follow-up calls, and nobody retypes a phone number from a clipboard.
If one of these costs you real money every month, describe it in four answers.
Thinking about what a process actually costs you?
The assessment is small, fixed-price, and tells you the real number, whether or not you ever build.
Worth a 30-minute conversation.