You should be able to ask your property data a question in plain English and get an answer in seconds, not export three reports and rebuild them in a spreadsheet. URBI does exactly that with HERO, an operator AI that answers questions and takes action from live building records.
Why does every answer live behind a report menu?
Because most property management software was designed to store data, not to answer questions. The data is in there. Getting it out means finding the right report in a menu of dozens, setting the filters, picking the date range, exporting to Excel, and then doing the actual thinking yourself.
Watch what happens when an owner asks a property manager something simple, like how maintenance costs trended against last year:
- Open the reporting module and hunt for the report that is closest to the question.
- Run it once for this year and once for last year, because the report was not built to compare periods.
- Export both to Excel and line the columns up by hand.
- Notice the categories do not match, fix them manually, and hope nothing got lost.
- Send the answer two days after the question was asked.
Every answer is an export away. The question took ten seconds to ask and an afternoon to answer. Multiply that across a portfolio and reporting quietly becomes one of the biggest time costs in the job. The report menu is not evil. It is just the wrong interface for questions, because a menu can only contain the questions someone predicted in advance.
What does asking your property data anything actually mean?
It means you type the question the way you would ask a colleague, and the software translates it into queries against live records, then returns an answer, a chart, or a completed action. No report menu, no export, no spreadsheet step in the middle.
Here is the concrete version. In URBI, a manager types "show me the occupancy rate for the last 12 months as a chart" and HERO builds that chart from live records on the spot. Not from last night's export. Not from a canned report someone configured months ago. From the same records the building runs on right now.
This only works when the AI sits inside the system of record. A general chatbot can summarize a document you paste into it, and we cover where that helps and where it fails in our piece on using ChatGPT for leases and bylaws. But a chatbot with no connection to your units, tickets, and payments cannot tell you which units are behind on fees, because it has never seen your building. The data has to live where the AI lives.
How does HERO understand a whole building?
HERO runs on a knowledge graph of the property: units, people, documents, service tickets, vendors, and more, stored as connected entities rather than isolated tables. That connection is the point. A vendor links to the tickets they worked, tickets link to units, units link to the people in them, and documents link to all of it. When you ask a question that crosses those lines, HERO can follow them.
To act on that graph, HERO uses tools in four categories:
- Retrieval: pull records, run queries, build charts and summaries.
- Operations: file service tickets, update work, schedule reminders.
- Booking: work with amenity and facility bookings.
- Communication: draft news posts and messages to go out through the platform.
Every tool call is written to a full audit log with its status, arguments, and result, and token usage is tracked per message. You can see exactly what the AI looked at, what it did, and what it cost. For anyone who has been asked by a board "how do we know what the AI is doing," that log is the answer.
Which questions become instant?
Questions that used to mean exports and cross referencing become a single sentence. Realistic examples a manager could type into HERO today:
- "Show me the occupancy rate for the last 12 months as a chart."
- "Which units are behind on condo fees this month, and how does that compare to last quarter?"
- "Which service tickets have been sitting in progress the longest?"
- "Which vendors handled plumbing tickets this year, and how many of those tickets are still open?"
- "Summarize the last board meeting minutes into five bullet points."
- "Which amenities get booked the most, and on which days?"
- "Draft a news post telling residents about the elevator maintenance on Friday."
- "File a service ticket for the leak in the parkade and set a reminder to check on it Thursday."
Notice the last two are not questions at all. That is deliberate, and it is the next point.
Can the AI act on the answer, or only report it?
HERO acts as well as answers. It files service tickets, drafts news posts, schedules reminders, summarizes long board minutes, and runs reports, all through the same logged tool calls. The pattern in practice is ask, see, act: you ask which tickets are stalled, see the list, and tell HERO to nudge the vendor or file the follow up, without switching screens.
For the bigger deliverables, HERO has a Chief of Staff mode that produces longer consolidated outputs, like a quarter end report or a yearly review draft, pulled together from the building's records. The Chief of Staff report is a fixed format today, not customizable, and we go deeper on what that means for boards and owners in our guide to AI owner and board reporting.
This matters most at portfolio scale. URBI supports multi property portfolios with per property permissions, so an operator running many buildings can ask questions building by building without stitching exports together. If that is your world, start with our guide to multi property management software, and see how this compares across the field in our review of the best AI property management software.
What are the honest limits?
HERO is the operator side, and it lives in the URBI web dashboard on desktop. It is available to any enabled user, including board members if the building has HERO turned on, but it is not in the resident mobile app. Residents do not talk to HERO at all. They talk to Arthur, URBI's resident facing AI, over phone, SMS, email, and in app chat in ten languages. The two are deliberately separate: Arthur serves the person in the unit, HERO serves the person running the building.
Two more things stated plainly. The Chief of Staff report is a fixed format today. And custom dashboards, where operators build their own live views and reports across every record, are on the roadmap, not shipped. Asking questions in plain English works now; pinning those answers to a dashboard you design yourself is coming. None of this replaces the manager's judgment either, a point we argue in full in will AI replace property managers. HERO removes the export and spreadsheet work between a question and its answer. Deciding what to do with the answer is still your job.
For residential buildings especially, this changes the rhythm of the work. The questions boards ask in meetings, the ones owners email about, the ones you ask yourself on a Monday morning, stop being report requests and start being sentences.
Frequently asked questions
Do I need to set up reports or know a query language first?
No. You type the question in plain English and HERO translates it into queries against the building's live records. There is no report builder to configure and no query language to learn. The tradeoff worth knowing: answers come from what URBI holds, so the more of your operation runs on the platform, the more questions it can answer.
Can residents ask questions about the building too?
Residents use Arthur, not HERO. Arthur is URBI's resident facing AI and it handles resident questions and requests over phone, SMS, email, and in app chat, in ten languages. HERO stays on the operator side of the platform, on the desktop web dashboard, working with the full record set that managers and boards are entitled to see.
How do I know what the AI actually did with my data?
Every HERO tool call lands in a full audit log with its status and result, and token usage is tracked per message. You can review exactly which records were retrieved and which actions were taken. This sits alongside URBI's permanent audit trail across the platform, covering actions, role changes, votes, payments, and AI tool calls.
If you are tired of treating every simple question like a reporting project, ask us the first one. Write to hello@myurbi.co and we will show you HERO answering questions from a live building.
