Almost exactly 3 years ago the world saw the culmination of years of collective research with the release of ChatGPT by OpenAI.

This marked a huge shift in the general public's relationship with Artificial Intelligence. Reaching 100 million users in two months.

Since then researchers, tinkerers and technologists (yes, not all nerds are the same) have pushed the frontier of what is possible with text-based reasoning.

We've begun using language as a mechanism for instructing computers to perform tasks. Today I'll explore the broad categories of solutions that have emerged for Property Management in the last 3 years.

I swear to God that I wrote and reviewed all of this with 0 ChatGPT slop writing, so forgive me in advance :)

AI for customer support

One of the first clear applications of AI was to bridge the legacy support systems or to empower them to be 10x better. Tools like Zendesk, Freshworks and Intercom all were forced to improve their capabilities to match smaller players. This was sufficient for people who want quick wins but for majority of property managers this wasn't sufficient to change day-to-day operations. Just another (sometimes expensive) tool for their offshore teams.

At times we're all guilty of using every knife in the kitchen to make the same meal we've eaten every Sunday for a decade. It can feel good to change our tools, but the measurable impact is often limited to the cutting itself… not the meal we're making.

AI for deal analysis

Critical AI companies (what we call "Foundational Model Providers" like OpenAI and Anthropic) began running hackathons and campaigns to explore the capabilities of these text-based systems.

Smart people very quickly realised that we can feed more data into these systems than a regular employee can manage at any given time.

How can this help property managers?

Well, deal analysis of course. As regional economies saw fluctuating interest rates managers across the world began adjusting their portfolio, some growing, some restructuring. Tools like Built AI emerged to help real estate investors including managers consider deal sourcing and screening.

Whilst this was tremendously valuable at points in time, for the traditional manager there are much larger inefficiencies that need plugging. The key to restructuring in times of distress isn't in just AI, it's in building a strong ship that can sail any waters.

AI for document processing & generation

It wasn't exactly sequential with this one, the same deal analysis opportunity that formed on the AI frontier also came hand-in-hand with document processing. AI vision models (the ability to view images using AI) unlocked a huge opportunity, becoming the largest room for gains in 2024. Managers could use basic tooling (like ChatGPT) to draft documents and executives often hire(d) consultants to optimise their most inefficient document flows.

Combining technical expertise and pipelines with existing document storage & human evaluation meant high upfront costs… but these were rapidly paid down within a year for many early adopters.

It was clear by the start of 2025 that we can begin using these AI systems to piece together fragmented aspects of property management. This posed the largest question that remains as I write this in Q4…

What's next? Where do we go from here?

We began building Opsansa (the only support operations solution that gives AI with personal memory to tenants) in 2025 after being some of the first in the world to build "agentic systems" and winning awards to prove our abilities; we even started the year building the UKs first AI rental agent and scaling to thousands of users.

Programmers globally quickly began building "agents for agents". "agents for managers" and shilling digital snake oil. With real estate being the worlds largest asset class it attracted the people building weak solutions. Through our experience speaking to property experts over the years we knew that management wasn't easy.

Property managers don't need AI as a knife in the kitchen, they need it to make a better meal… one that's quantifiably more efficient, that empowers the human experience and that is so good we don't need to call it AI.

Our approach in building Opsansa is to sit at the intersection of existing systems, customer support and employee brains.

Businesses need every piece of interaction data that makes a strong & scalable property management company.

In a decade you won't wish your tenants spent more time talking to Zendesk or calling your offshore support team…

You'll wish the data on the portfolio and your tenants was stored securely and processed insightfully so you can begin to automate all of your digital processes. A single domestic manager could handle thousands of units, having Opsansa teams handle rent roll, maintenance, personal reminders and everything else.

This is only possible by unifying interactions like acorns on a tree… with each tenant having their own acorn and your Opsansa instance being the tree. Acorns share the same nutrients and love but fundamentally are their own organism and… who knows… could eventually become their own tree?!

As AI advances and becomes more capable at task completion and interoperability (where we are now), we may see a world where tenants can self-manage with your executives oversight… empowering tenants to be close to the manager, to feel heard, to live at ease… without needing to smash their keyboard because they're sick of AI.

The last 3 years have taught us something important… it was never about "AI" — it was about how do we serve the people that make up our lives. Because if it isn't making the meal better then we don't need to spend hard earned dollars showing tenants new kitchen gadgets.

Thanks,
Rowan (Founder)