Building the AI Intelligence Layer for Institutional Real Estate

We’re excited to announce that Work-Bench has led MARC’s $5M Seed round alongside MXV Capital!

Institutional real estate manages trillions of dollars in assets. Owners have invested heavily in sophisticated technology to understand the revenue side of their properties: rents, occupancy, pricing, and market demand.

But there’s a glaring blind spot on the other side of the equation: operating expenses.

Despite all the sophistication around maximizing revenue, understanding and controlling what it costs to operate a building remains surprisingly manual.

That’s where MARC comes in.

MARC is building an AI-native spend intelligence platform that transforms fragmented vendor contracts, invoices, and operating data into actionable intelligence, helping institutional owners protect their net operating income (NOI).

And we believe the opportunity goes well beyond catching billing errors.

Why Now: Multifamily’s New Economic Reality

The industry has changed. Its operating infrastructure hasn’t.

Over the past two decades, multifamily real estate has undergone a fundamental transformation.

What was once a fragmented, locally operated industry has attracted enormous institutional capital. Large investment managers now oversee tens of thousands of apartment units across multiple markets, often relying on third-party property managers.

That model worked particularly well in an environment of rising rents, cheap capital, and appreciating property values.

But the economics have shifted.

Higher interest rates, rising insurance and labor costs, and pressure on valuations have forced owners to focus on operational efficiency. Meanwhile, property management fees are being compressed, leaving operators to do more with less.

When you can no longer rely on asset appreciation to drive returns, every dollar of NOI matters.

At a 5% cap rate, every $1 of recurring operating expense eliminated translates into $20 of additional property value.

Yet the technology supporting these increasingly sophisticated portfolios hasn’t kept pace.

A multifamily operator managing 25,000 units might have thousands of vendor contracts covering landscaping, elevators, security, maintenance, telecom, and dozens of other services.

Those agreements are scattered across emails, SharePoint folders, property management systems, and local offices, each generating invoices, amendments, renewals, and pricing escalations.

Accounts payable systems have become very good at processing invoices quickly. But they generally validate invoices against purchase orders or approved budgets, not the underlying contracts that govern what vendors are entitled to charge.

A vendor can increase its bill by 5%, remain within an approved tolerance, and get paid automatically - even if the contract permits only a 3% increase.

Multiply that across thousands of vendors and hundreds of thousands of invoices, and small discrepancies become significant financial leakage.

Historically, solving this required manually finding, reading, and reconciling enormous volumes of documents. The economics simply didn’t work.

AI changes that equation.

For the first time, it’s possible to turn unstructured contracts into structured, actionable data at scale and at a fraction of the historical cost.

That’s the opportunity MARC is going after.

The Product: From Processing Invoices to Understanding Them

AP systems were built for speed. MARC is building for intelligence.

MARC isn’t simply adding AI to an existing accounts payable workflow. It’s solving a more fundamental problem: the systems processing invoices lack the contractual context needed to determine whether those invoices are correct.

The platform connects documents and systems that have historically operated in silos, ingesting vendor contracts, extracting critical terms like pricing, escalators, renewals, and obligations, and transforming them into structured, searchable data.

Under the hood, MARC takes a “contracts as code” approach. Instead of treating contracts as static PDFs, it converts their terms into structured information that AI agents can reason over and act upon.

That unlocks three powerful capabilities:

1. Contract intelligence. A centralized, searchable understanding of vendor relationships, obligations, pricing, and renewals across entire portfolios.

2. Acquisition and disposition diligence. Analyze thousands of documents in hours instead of spending weeks assembling contract information when buying or selling properties.

3. Autonomous invoice validation. MARC is extending its platform to compare invoices against actual contract terms, flag overcharges, and catch discrepancies traditional AP systems miss.

Existing software helps companies pay invoices faster. MARC helps them understand whether they should be paying those amounts in the first place.

And the long-term opportunity is even more interesting.

The Data Moat: Understanding What a Building Should Actually Cost to Operate

Today, institutional owners have sophisticated data on what properties should earn.

But ask an asset manager what landscaping, security, elevator maintenance, or waste management should cost for a particular property in a particular market, and the answer often depends on local relationships, experience, and spreadsheets.

There’s no equivalent, comprehensive intelligence layer for operating expenses.

We believe MARC has the potential to build it.

Just as CoStar has built an enormous data asset around real estate markets, rents, and property values, MARC has the opportunity to build a similarly powerful intelligence layer for operating expenses.

As MARC processes millions of invoices and thousands of vendor contracts, it has the potential to build a proprietary, anonymized dataset of operating costs across properties, vendors, and geographies, with customers opting in to contribute their data.

And this wouldn’t just be a database of prices. By capturing the context behind each expense—property type, location, service scope, contract duration, volume, and escalators—MARC can enable meaningful comparisons across properties and markets.

Over time, this intelligence could help participating owners benchmark expenses, negotiate better agreements, consolidate vendors, and forecast operating costs more accurately.

Imagine evaluating an acquisition in a new market and knowing, before closing, whether a property’s operating expenses are above or below market. Or identifying opportunities to consolidate landscaping contracts across 20 properties at materially better pricing.

As more customers choose to participate, MARC’s anonymized spend intelligence has the potential to become increasingly valuable, with every new market adding another dimension to its insights.

That’s how MARC can evolve from contract intelligence into a foundational data and decision-making platform for institutional real estate.

Traction: Institutional Customers Are Already Leaning In

For an early-stage company selling into an industry known for long sales cycles, MARC’s initial traction is remarkable.

The platform already manages $100M+ in annual vendor spend across 30 US states and Canada, with customers representing approximately $90B in assets under management.

Customers include Magnolia Capital and Birge & Held, alongside some of North America’s largest institutional real estate operators. MARC was also profitable before raising this round.

One example from Birge & Held illustrates the value.

The firm operates approximately 25,000 multifamily units. MARC analyzed more than 50,000 invoice records from a single technology vendor, uncovering over 100 contract-backed overcharges. Work that would have required thousands of hours of manual review was completed in a fraction of the time.

But what stood out most during our customer diligence was how quickly customers began imagining additional uses for the platform.

One customer described finally gaining visibility into contracts scattered across property managers and internal systems for years. Another saw the opportunity to transform employees from data aggregators into decision-makers optimizing vendor relationships across markets. MARC is enabling customers to rethink how they manage their portfolios.

The Team: Relentless Execution

Some founders just refuse to slow down.

MARC is led by 24-year-old founder and CEO Aaron Devitt, whose entrepreneurial journey started when his family’s building lost heat for more than a week over Christmas and they couldn’t reach the property manager.

He set out to build better software connecting tenants and property managers, growing his first product to more than 40,000 monthly logins across the UK and Ireland.

But after spending time with institutional owners, Aaron recognized a much larger opportunity: helping them understand and control the enormous costs of operating their properties.

He expanded into North America and built MARC around that problem.

What has impressed us most is Aaron’s relentlessness. He’s constantly in front of customers, learning how they operate, finding creative ways to get into the room, and turning those conversations into product improvements and commercial momentum.

In an industry where enterprise sales cycles are traditionally measured in quarters, Aaron has demonstrated an ability to move extraordinarily quickly.

We love founders who combine enormous ambition with the willingness to do the unglamorous work required to win customers. Aaron embodies that mentality.

Why Work-Bench Is Excited

Three reasons we’re particularly excited about MARC:

A structural market shift. Institutional ownership, margin pressure, and growing operational complexity are creating an urgent need for a better way to manage real estate expenses.

A differentiated product architecture. MARC isn’t another dashboard layered on top of legacy AP software. Its contracts-as-code approach unlocks information existing systems were never designed to understand.

A compounding data advantage. By building a structured, hyperlocal dataset of vendor contracts and operating costs, MARC has the opportunity to create a proprietary intelligence layer that becomes more valuable as its footprint expands.

We believe these forces create an opportunity to build a category-defining enterprise software company.

And with Aaron and his team executing at the pace they are, we’re excited to be along for the ride.

We’re thrilled to partner with Aaron and the entire MARC team as they build the next generation of operating intelligence for institutional real estate.

Welcome to the Work-Bench family, MARC!

Let’s get to work. 🚀🚀🚀

You can read more about MARC’s seed round in the accompanying press coverage from This Week in Fintech.

The team is hiring for a Head of Engineering, learn more about the role and apply here.