Imperium builds villa management companies their own AI Operating System: the portfolio in one place, owner reports and rates prepared from your own data, and every guest answered, with a person approving the work.
Start the introduction →Imperium builds villa management companies their own AI Operating System, so the portfolio can grow without the office growing with it.
Bookings, costs and payouts sit in different systems, so what each property really earns stays unclear.
Reports go out late, and owners lose confidence.
Gap nights, and money left on the table.
Slow replies end up in the reviews.
Growth means more hires, and the margin shrinks.
When someone is off, things slip and the owner hears before you do.
Keep them. On their own, they stall for the same reasons.
Your systems, accounts and files aren’t wired in, so each answer only knows what someone pasted in.
That’s where your guests, owners and villa staff talk, so those conversations sit outside every answer.
What one person teaches it, nobody else gets. When someone leaves, it goes with them.
Someone still copies numbers between systems before every owner report.
“ChatGPT just answers certain questions. This builds stuff for you.”
We build the other two, around your business. Together, that’s your AI Operating System.
The brain. Claude, ChatGPT, or a local model on your own machines for sensitive work. When a better one comes out, we plug it in.
Your booking system, pricing tool, WhatsApp, accounts and files, connected and set out the way your business runs, so the answers are right.
Custom built for your company: your team, your systems, the way you work. On every desk and phone, your team asks in plain English and approves the work it drafts.
You own your data and everything built on it: infrastructure that grows with you. Swap the model or the app later and nothing is lost.
Occupancy, revenue and owner payouts across every villa, live.
Built and checked from your own data, every month.
Seasons and gap nights reviewed across the whole portfolio.
Every channel watched, and replies drafted.
Jobs logged, assigned and chased.
Accounts, payouts and costs in one picture.
The outcome for the average client.
Figures calculated from selected client interviews, across different companies and tasks. See the individual stories for context.
Read the figures and methodology ↗The tools differ, but the problems repeat: knowledge scattered across teams, reports that take days and work that waits on one person.
Our company and team experience spans property, hospitality, retail, technology and professional services.
Selected company and team experience, including earlier professional relationships.
Questions from the people doing the work shape the product: visibility, accountability, adoption and time returned to the team.
Live presentation · bringing AI into the operating conversation
Recorded conversation · questions before tools
Talking through where the work stalls
Listening to how the work really runs
Learning alongside business owners
Workshop · from questions to decisions
Sales, property management, daily rentals and land. Founder Sean Johnson, 22 years in real estate.
Seventeen villas under management heading for fifty, run by four people in the office. Admin was the ceiling.
The operating system over Hostaway, PriceLabs and Xero: owner reporting, revenue management and guest comms in one picture.
“Once the system is set up completely, 60% of time back, easily.”
Sean Johnson, Founder
Reports that took a staff member six hours take one.
Sean’s estimate once fully set up: 60% of the organisation’s time back.
The plan: 17 to 50 villas on the same office team.
Watch the film on YouTube ↗Sean’s client story and film ↗17 pool villas and a residential community, run by founder Shanny Poijes.
Tenders, purchase contracts, legal, accounting and marketing for a full development, with a plan to hire six to eight people to carry it.
The operating system as his back office: every document, number and message in one place. Contracts drafted by the system, quotes sent from his phone.
“I was thinking I might need six to eight people. Now I need one.”
Shanny Poijes, Founder
A planned team of six to eight became one hire.
A sale contract that cost $3,000 to $5,000 in legal drafting is now a button.
$15,000 a month saved, in his words: the people he no longer has to pay.
Watch the film on YouTube ↗Shanny’s client story and film ↗Shanny podcast, 17 villas, one employee ↗Property development and hospitality across several brands.
Several projects and venues moving at once, with each new one adding hires in marketing, finance and settlements.
One operating system across the whole team, with each role given the tools it needs.
“It was understanding what was giving us a return on our advertising expense, where people were working effectively, and getting the information through the organisation as well.”
Chad Egan, Founder & CEO
10 hires prevented in 60 days. Chad prices that work at hundreds of thousands of dollars.
The next hires are senior specialists. GeoNet won’t add fifty more in marketing, thirty in finance or forty in settlements, because the system covers that work.
COO Ben Barnes: some days it felt like a team of ten working for him.
Watch the film on YouTube ↗GeoNet’s story and the 60-day documentary ↗Ben Barnes, COO, GeoNet ↗Your own app, connected to the systems the portfolio already runs on.
Custom built for your company, on every desk and phone, in the dock next to ChatGPT and Claude.
Hostaway, PriceLabs, Airbnb, Booking.com, Xero, WhatsApp, Gmail and Drive, chosen from a catalogue of 1,400+ tools. Connect only the ones you authorise.
Your portfolio’s knowledge, connected behind every answer and action.
Guest, owner and staff conversations, connected to the same data. Replies drafted for approval.
OpenAI and Claude models, with local models on your own machines for sensitive work.
Owner reports, rates and replies are prepared for review. You set who approves what.
Occupancy and rates across every property, owner reports ready to approve, maintenance being chased and every guest answered.
You ask in plain English.
It checks PriceLabs, Hostaway and Xero.
It answers from your own data, and shows where it came from.
It drafts the next step. A person approves it.
Your own app, in the dock next to ChatGPT and Claude.
The few decisions worth your time, with the impact and the evidence next to each one.
Each number traces back to the system and the rows it came from, so your team can check it.
Emails, messages and changes are drafted for a person to approve. You set who approves what.
Every installation is engineered inside the business, around the systems and people already running it.
Structure company information so context can be found, understood and traced to its source. Taxonomy and retrieval make the knowledge useful beyond a single conversation.
Connect the data, tools and permissions behind daily operations. Each installation has to reflect how the company actually works, from the source of a record to the team responsible for using it and reviewing the next action.
Work with sales, marketing, finance and operations to decide what matters, what should change and how the team will measure progress. A defined workflow, an accountable owner and a useful baseline give the installation a practical starting point.
An unpaid owner payout belongs to a villa, an owner, a booking and the person responsible for clearing it.
The AI OS keeps those relationships together, so every answer starts from the business itself rather than an isolated prompt.
The engineers who build it
stay with it once your
team is using it.
Map the bottlenecks
in the business.
Wire in the systems
the portfolio runs on.
Put your team on the app,
then add the next areas.
Choose an internal champion and a real workflow, such as the monthly owner report. Establish what good looks like, then work through the difficult cases together.
The goal is capability inside the business: people who understand the OS, trust its boundaries and can use it without waiting for the founder.
See how in-house capability is built ↗“Reduce the time we spend on manual procedures and workflows and filling out spreadsheets…”
A 2025 Project NANDA report found that most organisations studied had not demonstrated measurable business returns from GenAI.
Read the preliminary research ↗The underlying model will change. The company’s knowledge, definitions, workflows and permissions are the foundation the team keeps building on.
Every installation trains internal champions: people who learn the OS, test it against real work and help the rest of the company adopt it.
Explore the approach ↗The product takes shape alongside the teams who use it. We map the company’s questions, build around its priorities and stay close as the team puts the OS to use.
Product engineering and commercial operations meet in the same responsibility: understanding the company, building around its priorities and helping the team put the OS to work. The founders remain accountable for the quality of that relationship.
Chief Executive Officer
Leads company direction and product, drawing on software, data procurement and resource-planning experience. His focus is the architecture behind ARX and the engineering standard required to make company knowledge useful in daily work.
Chief Operating Officer
Leads operations, installation and customer success, drawing on the analysis of marketing, sales and operating performance. His focus is a clear business problem, useful measures and the internal ownership needed to make the OS part of everyday work.
Meet the founders ↗Working session · mapping the company before the build
Architecture, product engineering, security, design and delivery.
Eddie Wong
Chief Architect
Nirav Shah
Lead Product Engineer
Otto Bonilla
Senior Software Engineer
Yuri Ivanov
Chief Security Officer
Amjad Ajzannay
Product Experience Lead
Bailey Spitzer
Creative Director
Our background combines software, data procurement, migration and resource-planning systems for large organisations with the analysis of marketing, sales and operations. That experience informs how we structure company knowledge, connect its systems and improve the work behind the numbers.
An AI OS earns its place through the work it helps people do. These principles guide how Imperium builds and runs every installation.
Learn the tools, responsibilities and daily work before configuring anything.
Make company context available for review alongside the answer.
Set connections and permissions around the people who use them.
Install alongside the team and develop internal ownership.
Keep engineering close as the business and its workflows change.
On stage · AI operating systems for business
Working sessions turn questions into product requirements. Conversations keep the engineering grounded in the company’s reality.
Time together, beyond the screen
Company workshop · space to work together
Conversations with operators
GeoNet · meeting the leadership team
Applications built on expert knowledge, structured company information, taxonomy and retrieval, search by meaning.
One view of company knowledge, shared business definitions, clear workflow ownership, operational partnerships.
Continuum Recruitment and Core Concept Living. Answers that show their sources, company dashboards, retrieval and memory.
Each company’s own app, approvals for connected actions, team onboarding, a catalogue of 1,400+ tools and sharing across the team.
Shared work that carries the company’s context.
Development themes by quarter, not exact release dates. October 2025 to April 2026 is the company’s account; May to September 2026 comes from dated records.
View development evidence ↗As models improve, the business’s organised data and workflows become more valuable.
It goes in stages. We connect and organise your data first, the bigger job, then put your team on the app. GeoNet was live across its team in 60 days.
We build it, train one person inside your business to run it, and stay on for support and the next areas you add. How much ongoing help you want is your call.
Often, work your own team could own. One person with the system can run what you paid a marketing agency for, and everything it learns stays in your business.
Models will keep changing. Your data, organised, is the infrastructure for the next ten years. Plug in a better model and nothing your business has built is lost.
It isn’t moved. Your data stays in your own systems and accounts. When someone asks a question, only what’s needed to answer it goes to the AI model.
You don’t give it to us. We never hold a copy. Sensitive work can run on a local model on your own machines.
You own your data, files and setup: everything your business has taught it. We own the app you see it through. Leave, point Claude or ChatGPT at the same files and carry on.
“I can’t have one of my staff asking for the financial report and it just spits it out.”
Agreed. Access is set by role. Sales sees the pipeline, finance sees finance, and you see everything.
Read the security overview ↗Grow the portfolio without growing the office.
A 45-minute audit call with the people who run the business.
We map the bottlenecks in your business, then come back with a roadmap: where we’d start, and what it takes.
Worst case, you keep the roadmap.
Book a 45-minute call ↗Or reply to whoever sent you this link.
Every client story and film ↗
