Brian Connolly, Founder + CEO, Feasibly
Brian Connolly is the Founder and CEO of Feasibly, an AI platform transforming how commercial real estate professionals conduct market studies and financial feasibility analysis.
Connolly spent years conducting market and financial feasibility work at Victus Advisors and is now focused on applying agentic AI to these real-world commercial real estate workflows. His goal is to help developers, investors and lenders dramatically reduce the time and cost associated with early-stage deal evaluation through smarter AI deployment.
In the following Q&A, Connolly shares why the industry has struggled to extract value from generative AI, why generic chatbots have fallen short of their revolutionary promise, and how a new generation of “agentic” AI systems is expected to deliver measurable business impact. He also discusses how these tools can bring efficiency, reduce risk across the industry, and accelerate the path from idea to viable project by providing reliable insights at critical milestones.
There’s been significant investment in AI across commercial real estate. Why hasn’t it translated into meaningful results?
The core issue is that most commercial real estate organizations started with the wrong tool. They invested heavily, upwards of millions of dollars for some, in generic AI chatbots or copilots. They had the expectation that these investments would support decision-making and automate workflows and routine tasks.
However, those systems weren’t designed for the complexity of CRE and can’t manage nuanced workflows. They regularly provide a different answer to each employee who asks the same question.
A 2025 MIT study found that 95% of organizations reported no measurable P&L impact from their generative AI investments, despite enterprise spending of $30-$40 billion on the technology.
These chat-interface AI systems lack context and struggle with multi-step processes, so instead of driving efficiency, they introduce uncertainty. This is why commercial real estate companies aren’t seeing any measurable impact: it isn’t so much that the technology doesn’t work, the technology just isn’t being applied correctly.
What’s fundamentally broken about the way commercial real estate companies are approaching AI today?
It comes down to a misalignment between broad, generalized AI models and highly specific business problems. The commercial real estate industry is a data-heavy and process-driven industry. You can’t just point a general-purpose AI model at a massive dataset and hope it produces reliable and repeatable outcomes.
Without structure, domain expertise and clearly defined workflows, AI can only exist as a novelty. Despite 88% of investors, owners and landlords running AI pilots, 60% report being unprepared to effectively implement AI into business workflows, according to JLL’s survey of over 500 senior decision-makers last year. Many pilots simply stall out and don’t move past the experimentation phase.
You advocate for moving beyond chatbots to “agentic AI.” What does that mean in practice?
Agentic AI involves building systems that are designed to do specific jobs, unlike chatbots that just answer questions. These AI agent systems are structured, goal-oriented and capable of executing multi-step workflows with consistency and minimal human oversight.
Agentic AI systems can be trained and deployed for highly targeted use cases within an industry. While chatbots aimlessly scour through a company’s sprawling collection of data, agentic AI systems are purpose-built to act as precision instruments. They are directed to find key information in precise locations and instructed on how to apply it for specific business use cases.
It’s the difference between an intern searching through files for answers and a trained specialist who already knows where to look, what matters, and how to act on it.
Why does that distinction matter so much in commercial real estate?
Consistency and accuracy are extremely important in the commercial real estate industry. When you’re evaluating a deal, underwriting risk or making a capital allocation decision, navigating variability is a critical step toward any successful project.
Chatbots create additional uncertainty because they generate responses dynamically, whereas agentic systems remove variability by following structured processes. That makes them far more reliable for real-world business applications, especially in cases where precision matters.
Where are you seeing the most immediate impact from this approach?
Market studies and financial feasibility analysis are compelling examples of the power of purpose-built AI solutions. Traditional feasibility studies for a commercial real estate project can cost $50,000 or more and require months of human analyst time.
Commercial real estate developers, investors, and lenders rely on market feasibility analysis to assess whether a project makes financial sense, but the slow and expensive nature of the process can prove an insurmountable roadblock for many executive decision makers. Agentic AI has fundamentally transformed this process, reducing the time and cost involved by up to 90%.
What’s enabling that level of speed and efficiency behind the scenes?
The orchestration of multiple specialized AI agents, where each agent is responsible for a discrete part of the workflow, allows for increased speed behind the scenes. In Feasibly’s system, for example, tasks such as data retrieval, market validation, benchmarking, forecasting, and narrative synthesis each have a dedicated, customized AI agent.
By breaking that process into defined steps and then assigning each step to a purpose-built agent, what was previously a highly manual process is transformed to one that is automated and efficient. It’s not entirely hands off, however. Keeping a human expert in the loop is still essential to validate outputs, perform quality control, and ensure compliance with industry best practices.
How does this change the role of human analysts?
It elevates the analyst role. Instead of spending weeks gathering data and then assembling reports, analysts can focus on interpreting the data, judging it, and creating strategies. By automating workflows that traditionally require extensive manual effort, it significantly reduces turnaround time. A human analyst is always in the loop to confirm results, ensuring accuracy and compliance.
What broader industry challenges does this help solve?
Using AI to automate market studies solves a growing challenge in real estate development. Mixed-use projects are becoming increasingly common, often with a public-private partnership component, demanding more complex analysis. At the same time, a surge of first-time developers and private investors are planning smaller projects for which they need quick, reliable market and financial analysis to test ideas and secure early funding.
Agentic AI makes professional-grade feasibility analysis accessible to more builders, funders and planners, which helps projects move forward when they would otherwise stagnate or be shelved. Lowering the barrier of entry for this important stage of real estate development could have significant implications for commercial real estate markets, from multifamily housing to retail and hospitality. Notably, making it cheaper and easier for developers to assess the viability of housing projects could help alleviate the nation’s 600,000-unit apartment shortage.
What kind of ripple effects could that have across commercial real estate?
Agentic AI applications have the potential to unlock a significant amount of stalled or unrealized development. When market and financial feasibility analysis becomes faster and more accessible, more ideas can be tested and validated earlier in the process.
Across sectors, particularly in housing, developers can more easily assess project viability to accelerate the delivery of new units and address supply constraints.
What should commercial real estate leaders be doing right now to rethink their AI strategy?
Commercial real estate leaders need to change their mindset and shift from experimenting with AI to applying agentic systems that execute. Instead of asking, “How do we use AI?” they should be asking, “What specific problems are we trying to solve by using AI?”
From there, leaders need to adopt purpose-built solutions that are designed to solve those specific problems. Generic tools might be easy to deploy, but they rarely deliver measurable bottom-line outcomes. The companies that leverage targeted, agentic systems will be the ones that see real competitive advantage with their AI strategy.