putilov_denis - stock.adobe.com

Fintech vendor releases AI tool for commercial real estate

Unlike the residential real estate market, growth in commercial real estate has been slow. A fintech vendor is seeking to fill a market need for commercial real estate data.

Enterprises operating in the commercial real estate world now have another AI tool they can use to determine the valuation of different properties.

On Jan. 31, fintech vendor Aqueous Asset Inc. introduced its Ava valuation platform, which uses AI software to enable users to make data-informed real estate investing decisions.

Users can type in the address of any multifamily property in the U.S. and get valuation data for property, projected rent rolls, and supporting sales and lease comparative data. Ava pulls data from seven different APIs, including Zillow, Realist and Moody’s Analytics.

The real estate AI system is designed to combat the slow growth of the commercial real estate market during the COVID-19 pandemic, according to Aqueous.

In this Q&A, Nick Segal, CEO of Aqueous, explains how the vendor's technology compares with other real estate tech systems.

What makes Ava different from other real estate tools that use AI?

Nick Segal, CEO, Aqueous Asset Inc.Nick Segal

Nick Segal: Focusing on the commercial space, it's very different because we've all become accustomed to the Zillow valuation tool, and Redfin has one. There are several automated valuation mechanisms. In the commercial space, it's far more difficult. Our analogy is it's like a block of Swiss cheese. It has data sources, and we use seven different API data sources. But each one doesn't give you the complete picture.

Part of Ava's intelligence is the ability to dissect all the different information sources in real time to fill the holes, and then present you with a head start so that you can then refine the value. Our tool is very much human interactive. It's an opportunity for you to refine each data point so that you can come up with the most accurate and credible valuation possible. So many of the tools now, they spit out their number, or they'll spit out a range, and you have no ability to interact with it. The interaction capabilities are one of our definite distinguishing prospects.

Commercial developers like having control of that [real estate pricing] information, and they're hard-pressed just to give it away because that can be a competitive advantage.
Nick SegalCEO, Aqueous Asset Inc.

Why is AI in commercial real estate so hard?

Segal: It's a daunting task. They like it that way. Commercial developers like having control of that [real estate pricing] information, and they're hard-pressed just to give it away because that can be a competitive advantage. For us to really democratize the information and make it accessible in a SaaS product where you can buy a subscription monthly, you get access to not only all the API feeds, but also Ava's artificial intelligence and machine learning, and then refine it all in real time.

In real estate, and in AI, there's room for bias, especially racial bias when it involves neighborhoods and properties. How do you make sure your tool doesn't further that bias?

Segal: Commercial real estate starts with your income. The valuations are defined on how the building performs. So it's very difficult to have a bias toward valuation when you have public access to current rents for studios, one bedroom, two bedrooms, three bathrooms. Based on the improvement: How new is the building? How old is the building? Supply and demand drives rent cost. So that creates a pretty level playing field.

Where you then can bias the information, which we will always have, is which comparable values do you use to determine the value of the separate property, right? If I had 10 comps to choose from, and I want to skew it to my advantage, I can take the comps that are within the geographic area, but there could be better comps than the columns that I could be using, which could then inflate or depress values.

How do you plan to make this AI tool better?

Segal: What we want to do is expand the intelligence. Right now, we're driven by algorithmic equations, right? But every time someone types in what their current rent roll is across the country, it captures that information. So we have a valuation. We then can start to rely more heavily on the actual valuation as opposed to algorithmic valuations that were derived from the API. Our intention is just to get smarter and smarter and more authentic.

In terms of 'are we trying to exclude or alienate anybody,' no, we want this to be as inclusionary as possible, because that transparency creates a more level playing field, and people don't feel duped. Commercial real estate, just like residential real estate, is a big investment, and if you don't have all the information to feel confident that you're making an informed decision, you're going to either second-guess yourself or you're going to potentially let fear stand in the way of you actually doing what could be a very good deal -- or [you might be] getting into a deal that you in retrospect go, 'Why in God's name did I invest in this deal?'

Editor's note: This interview has been edited for clarity and conciseness.

Next Steps

How to balance data access and security in fintech testing

Dig Deeper on Artificial intelligence platforms

Business Analytics
Data Management