May 30, 2023
Artificial intelligence holds promise in almost every industry. But in commercial real estate, one of the world's largest and most data-rich business sectors, there are still questions about AI's data-processing capabilities and accuracy. Could an AI adoption model that blends human due diligence with AI's processing power be the ticket to moving forward with widespread automation in an industry like CRE? Philip Russo of the Commercial Observer talks at length with Prophia CEO, Cameron Steele, about the way forward in spite of this limitation and how Prophia is currently achieving 100% accuracy with its data abstraction model.
Efforts to digitize and streamline the real estate industry face challenges despite advancements in artificial intelligence (AI) and machine learning. While AI technologies like generative AI, which can analyze and interpret complex real estate documents, offer benefits such as speed and efficiency, they are not a complete substitute for human expertise. In fact our own AI achieves its 100% accuracy rate deploying a combination of human and computer-generated data synthesis to provide CRE professionals with one of the most advanced automation models currently on the market.
That’s not to say however that CRE doesn’t pose challenges for AI. In a recent piece written by Philip Russo in the Commercial Observer, Philip describes some of the limitations generative AI models like ChatGPT face in industries like CRE where mountains of data can exist in a single portfolio alone. It’s a tricky balancing act to achieve; toeing the line between data accuracy and processing power. But the secret really could lie within the people at the heart of the tech.
“Every tenant we onboard on behalf of our customers has about seven documents and about two-thirds of all of the annotations we do are through auto annotation,” Prophia CEO, Cameron Steele, explained. “The remaining third, we’ll do both technical audits as well as human review to make sure the existing annotations are done accurately. That takes us about 30 to 40 minutes per tenant today. The bigger burden on us is the review, which is harder to build tools around, but we’re working on that.”
But the human element remains critically important to processing and synthesizing nuance in commercial leases—which are all unique and complex in their own right. Thus, it seems the answer to successfully integrating AI into a document-rich industry like CRE is keeping the connection with the human element, relying on experts to enhance the machine and vice versa.
Luckily for Prophia, our model’s training data dates back to 2018 when we began building our first client relationships. As it currently stands, Prophia’s training dataset may be the most sophisticated and rounded in the industry, making our proprietary ML algorithm particularly powerful and accurate, but of course we strive to develop the model’s intelligence further.
Philip concludes that, although AI and machine learning technologies have improved document abstraction and analysis in the real estate industry, achieving complete accuracy remains a challenge. Human expertise and domain knowledge continue to play crucial roles in effectively interpreting and understanding real estate documents, making a fully automated solution elusive at present.
Read the full piece by Philip Russo on Commercial Observer.