AI property management assistants: managing tenant and landlord communication

What an AI property management assistant does, how it handles tenant and landlord communication, where it falls short, and how to choose one for your portfolio.

A property manager’s day is shaped less by the properties than by the people around them. Tenants ring about a leaking tap, applicants want to know if the unit is still available, owners email to ask why the rent hasn’t arrived, and a tradesperson needs access details by lunchtime. An AI property management assistant is designed to take on the routine part of that communication: answering, sorting and following up, so the property manager can spend their time on the conversations that genuinely need them.

This guide explains what these assistants actually do, how they handle the different people a property manager deals with, where they fall short, and what to look for if you’re weighing one up. It’s written for property managers and principals in Australia, New Zealand and the United States, where the job is much the same even if the words differ: inspections or tours, strata or HOA, rental enquiries or leasing leads.

What an AI property management assistant is

An AI property management assistant is software that holds conversations with tenants, owners, applicants and trades by phone, chat, text and email, and takes action on what it hears. It differs from the scripted chat widgets many agencies tried a few years ago in two important ways. First, it understands what someone is asking in their own words rather than forcing them through a menu. Second, it can do something with the answer: log a maintenance request, book an inspection, update a record or pass the conversation to the right person with the details already collected.

The term covers a wide range of products. Some property management software now includes built-in assistants that draft replies or summarise messages for staff. Others are front-line agents that answer the phone and website on the agency’s behalf. Some are tools you configure yourself, and others are managed services that set up and run the assistant for you. Knowing which kind you’re looking at matters, because the work involved, and the results, are very different.

How it handles each person in a property manager’s week

Tenants

Most tenant contact falls into a handful of categories: maintenance, rent and account questions, lease questions, inspections and moving in or out. An assistant can answer the routine questions from information the agency approves, such as how to pay rent, what to do about a blocked drain or when the next routine inspection is due. For maintenance, its job is to capture the problem properly the first time: the property, what’s happening, how long it has been happening, photos if they help and access details. It then applies the agency’s rules about urgency, so that a burst pipe reaches the on-call person straight away while a sticking door handle is logged for the next business day. For more on this part of the job, see our practical guide to AI for property management.

Landlords and owners

Owners rarely leave an agency because of one mistake. They leave because of silence: the repair they heard about late, the question that took three days to answer. An assistant can answer factual owner questions where the answer is already in your system, such as whether the plumber has attended or when the inspection report will be ready, and keep owners updated at the moments that matter. What it shouldn’t do is give advice on the agency’s behalf. Rent recommendations, disputes and decisions about major works belong with the property manager. We cover this in more detail in keeping landlords informed.

Applicants and prospective tenants

Leasing enquiries arrive when listings go live and often in the evening, when people are browsing. The assistant answers questions about availability, rent, pets and parking from the listing details, books inspections or tours, and follows up with people who inspected but haven’t applied. Decisions about applications always stay with the team. Our article on AI for rental enquiries goes deeper on this.

Trades and contractors

Trades usually need access details, approval to proceed or confirmation that a job is complete. An assistant can pass on approved access instructions, record that a job has been finished and prompt the property manager when an approval is needed, rather than leaving a voicemail trail.

What an assistant shouldn’t do

The most useful assistants are clear about their limits. They shouldn’t make decisions that need judgement, such as approving an application, agreeing to a rent reduction or deciding who is at fault for damage. They shouldn’t give legal advice about tenancy rights. And they shouldn’t pretend to be a person. Customers are understandably wary of AI in customer service: in a Gartner survey of 5,728 customers, 64% said they would prefer companies didn’t use AI for customer service, and the most common concern was that it would make it harder to reach a person (Gartner, 2024). The practical answer is an assistant that says it’s AI, makes it easy to reach a person and passes the conversation over with everything already collected.

The lineWhat the assistant does, and what stays with people

A good assistant

  • Answers routine questions from approved information
  • Captures maintenance properly and applies your urgency rules
  • Books inspections and tours
  • Says it’s AI and hands over to a person easily

What stays with your team

  • Approving applications
  • Rent reductions and arrears decisions
  • Disputes and complaints
  • Advice to owners

Rules differ by country, so the assistant needs your rules

Property management rules vary between Australia, New Zealand and the United States, and within each country by state or city. Urgent repair rules are a good example. In New South Wales, the government lists repairs such as a burst water service, a gas leak, a dangerous electrical fault and a broken hot water service as urgent, and tenants can recover up to $1,000 for urgent repairs they arrange themselves if they can’t reach the landlord or agent (NSW Government, 2026). In New Zealand, a landlord must reimburse a tenant for urgent repair work if the tenant made reasonable attempts to let the landlord know first (Tenancy Services, 2026). In the US, repair obligations are set by state and local law.

The lesson for choosing an assistant is simple: it needs to follow your agency’s rules for your jurisdiction, not a generic script. That means agreeing in advance what counts as urgent, who gets alerted and what tenants are told, and checking that the assistant applies those rules consistently.

How to choose an AI property management assistant

Start with the problem you most want to solve. For most agencies it’s one of three things: calls and messages going unanswered, especially after hours; maintenance requests arriving in a mess and getting lost; or follow-up with applicants and owners slipping when the team is busy. An assistant that solves your biggest problem well is worth more than one that does everything adequately.

Then ask practical questions. Does it work with the systems you already use, or will someone have to copy information across? Can you see every conversation in one place? Can your team step into any conversation? Who sets it up, who checks how it’s performing and who improves it? Some assistants are platforms you configure and maintain yourself, which suits agencies with the time and skills to do it. Others are managed for you. Our comparison of DIY platforms and custom AI agents covers the trade-offs.

Finally, ask how the assistant treats people. It should say it’s AI, answer politely and briefly, and make it easy to reach a person at any time. If it doesn’t, the time it saves will be cancelled out by the complaints it creates.

How Shift AI runs tenant and owner communication for property managers

Shift AI isn’t a tool your team has to set up and maintain. We set up and manage the assistant around how your agency already works, and every call, chat, text, email and form is answered, sorted into a sales enquiry or a support request, and followed through to a result. That model has three connected parts.

Run: AI agents set up around your portfolio

Our AI agents answer tenants, owners, applicants and trades by phone, chat, text and email, day and night. They’re configured around your properties, your approved information and your rules: what counts as urgent, who gets alerted, which questions they can answer and which go to a person. They log maintenance requests with the details your team needs, book inspections and tours into your calendar, and update records in the tools you already use, including AppFolio, HubSpot, Google Calendar and Microsoft Calendar.

Scale: people for the conversations that need judgement

Some conversations need a person: an upset tenant, a dispute, an owner deciding about major works. You choose who handles those, part by part. It can be your own team, who can join any conversation in the Shift app, or a Shift specialist we manage, who follows up on your behalf using the same history the AI collected. The customer doesn’t start again, because the context comes with the conversation.

Management: reviewed and improved every month

Every month we review how things are going with you: what was answered, what was escalated and why, how quickly requests were picked up and where the assistant needs better information. Your scorecard shows the numbers, and we adjust the assistant’s knowledge and rules based on what we find.

In practice, a tenant who rings at 8pm about no hot water is answered straight away. The AI captures the details, applies your urgency rule, alerts your on-call contact and tells the tenant what happens next. In the morning, the request is in the Support queue with a full record, the owner has been updated in line with your process and the property manager simply confirms the trade.

If your team spends more time answering the same questions than looking after owners and tenants, book a free AI strategy sprint. In 30 minutes we map how calls and messages reach you today and where they slip through, then show what an assistant would handle and what would stay with your team.

Free AI strategy sprint

Let's get your customers handled.

In 30 minutes, we'll look at how your customers reach you today, show you where they slip through, and map the right mix for your business. Then we look after the rest.

Free setupLive in two weeksPay once it's live and working