AI for property managers does not have to mean expensive tools, complicated setups or technical knowledge. Even a small short-term rental business can start with something very simple: organizing what it already knows about its properties and letting artificial intelligence find the right answers.
What is the WiFi password? What time is check-in? Who do we call if the air conditioning breaks down? Does an owner’s approval need to be obtained for a repair and, if so, above what amount? As a portfolio grows, such simple pieces of information often end up scattered across emails, files in Google Drive, messages on WhatsApp and –most importantly– in the mind of the property manager.
A simple solution to this problem was presented by Uvika Wahi of Rental Scale-Up in a masterclass by RSU by PriceLabs and RevLabs by PriceLabs: creating a “property brain” with Google’s NotebookLM. In practice, this is a knowledge base where the manager adds the documents they already use for their properties and can then search for the information they need simply by asking questions to AI.
The important thing for a small property manager is that they do not need to set up a complex AI system or create a new database from scratch. They can start with the files they already have – check-in instructions, house manuals, rules, technician details and maintenance procedures – and turn them into a shared “second brain” for their team.
AI for property managers: From “ask the manager” to “ask the property brain”
In many management companies, a large part of the operational knowledge is concentrated in a few people.
A new team member wants to learn what applies to late check-out. Someone from guest support needs to find which technician services a specific property. An employee covering the night shift needs to know what to do if the air conditioning stops working.
The usual solution is a message or a phone call to the person who “knows.” The problem is that this process is difficult to scale as a company grows.
With NotebookLM, the manager can gather the documents relating to one or more properties and the team can then search for the information through questions in natural language.
Instead, for example, of searching through different files to find out what applies in the event of a breakdown, someone can simply ask: “What do I do if the air conditioning is not working at this particular property?”
Why it matters where the AI “learns” from
One of the key advantages of NotebookLM is that it works on the sources entered by the user.
This is particularly important in property management, where the correct answer is not necessarily the generally correct answer.
For example, a guest may request early check-in. The policy, however, differs from property to property. At one it may be allowed free of charge, at another it may be charged, and at a third it may not be allowed at all.
The same applies to maintenance, owner approvals, communication hours, suppliers or emergency procedures.
In the demonstration presented at the masterclass, when a question was asked about information that was not included in the documents –which is the nearest airport– NotebookLM did not attempt to fill the gap with an arbitrary answer, but stated that the information was not available in the sources.
This is precisely what makes such a system useful as an internal tool for a company: the team knows that the answers are based on information that has been approved and stored by the business itself.
What a “property brain” can include
The initial knowledge base does not need to be complicated.
It can start with the documents the property manager already has: property guides, house manuals, checklists, agreements with owners and supplier details.
Among other things, the following can be gathered:
- the basic characteristics of each property,
- check-in and check-out instructions,
- WiFi passwords and access instructions,
- house rules,
- policies for early check-in and late check-out,
- details of partners and technicians,
- procedures for breakdowns and emergencies,
- spending limits that require owner approval,
- the internal procedures followed by the team.
The files can come from Google Drive or be uploaded directly to NotebookLM.
The critical point, of course, is the quality of these documents. If the procedures are outdated, contradictory or incomplete, AI cannot correct the poor organization of information on its own.
In this sense, creating a “property brain” also functions as an exercise in organization for the company itself.
The example of the broken air conditioning
A characteristic example was presented at the masterclass.
The question to NotebookLM was what the team should do if the air conditioning breaks down.
Instead of a general answer, the system searched for the information in the property documents and provided a specific procedure: initial checks on the thermostat, the electrical panel and the batteries and, if the problem remained, the details of the specific partner who should be notified.
At the same time, it also stated what the employee should not do: contact the owner directly, as the specific issue was covered by a maintenance partner.
In another question, regarding a repair above $300, the system identified from the documents that above this amount the owner’s approval was required and provided the relevant communication procedure.
For a team managing dozens or hundreds of properties, this exact detail can make the difference: not simply finding an answer quickly, but finding the right answer for the specific property.
The greatest usefulness may be in onboarding
The application is not limited to serving guests.
One of the most obvious benefits lies in onboarding new employees and partners.
Instead of a new employee needing weeks to learn which supplier services which property, which exceptions apply to each property and where the different procedures are located, they can have access to a single knowledge base.
This is particularly important for companies with seasonal staff, shifts or external partners, where the same people are not always available.
The notebook can be shared with team members, so that everyone has access to the same information.
One notebook or many?
For a small portfolio, a manager can gather more properties in one notebook.
As the company grows, however, it may be more functional to organize the knowledge differently: one notebook per property, per area, per building or even per function, such as guest support, maintenance and owner relations.
There is no single correct structure. The logic is to create a system in which the information remains easy to update and, above all, reliable.
From SOPs to AI knowledge systems
The idea is not entirely new. Organized property management companies have been using SOPs –standard operating procedures–, manuals and internal knowledge bases for years.
The difference is that until now, the employee had to know where the right document was, open it and find the information they needed inside it.
With AI, the process is reversed: the employee asks the question and the system locates the relevant point within the available sources.
Some property management systems and other platforms in the industry already integrate similar capabilities. NotebookLM, however, offers a relatively simple way for a company to test this approach using the documents it already has.
And perhaps this is the most important point for property managers: AI does not necessarily need to replace a PMS or take over the entire communication with the guest to have practical value.
It can start with something much simpler: stopping the same question from being answered ten times by ten different people.
