Meet the Architecture Firm That Runs on AI

Artificial intelligence is rapidly becoming a core component of our daily lives, and architecture is emerging as one of its most promising frontiers. But what does AI really mean for the design industry? Could it lead to projects being generated at the push of a button?

Cove, an Atlanta-based architecture firm, is the first design studio to be powered by this technology. The company combines data-driven tools with advanced simulation, helping architects make more informed decisions, optimize workflows and deliver sustainable, high-performance buildings.

Recently, the firm launched Cove Architecture, its full-service architecture practice propelled by a proprietary two-part AI framework that has been in development for more than a decade and required more than $25 million in R&D investment.

Multi-Housing News sat down with Co-Founders Sandeep Ahuja and Patrick Chopson to find out all about cove’s inception and the duo’s shared vision for how AI will shape the future of architectural practice and design innovation.

---

READ ALSO: 5 Ways AI Tools Advance Multifamily

---

First, tell us more about your first days as a team. How did you two meet?

Ahuja: We met at Georgia Tech and have been working together for nearly 15 years.

Chopson: We first teamed up at Perkins & Will after school. One day Sandy said, ‘Let’s start a consulting practice for building science,’ and I was in. We started our first company together, later got married and then raised over $36 million to build the cove technology company.

What would you say was the inspiration to start cove and how has your vision evolved over the years?

Ahuja: What inspired me to start cove was the stagnancy in the built environment. Buildings account for about 39 percent of global carbon emissions and many industry workflows are outdated. With my background in building science and architecture—and growing up in diverse countries like Saudi Arabia, Russia and Tanzania—I felt a deep drive to create meaningful change.

Chopson: I’m an architect’s architect—I love design, but I’m also passionate about the technical side. For me, architecture is the intersection of art and rational systems.

As fighting climate change became an important issue, initially, I saw this as a way to create more interesting aesthetics. Over time, I realized it’s more than that: It’s also about preserving the profession by enabling it to adapt to an increasingly data-rich world. Technology allows us to manage complexity and maintain design excellence, and that’s where my passion currently lies.

Cove follows an AI-driven, human-centered approach. How do you balance automation with the creative and human aspects of architecture?

Ahuja: Technology is great at handling the non-sexy, repetitive and data-heavy parts of the design process. That frees up designers to focus on creativity.

For example, I spoke to a principal recently who said junior architects still redraw trash chute details from scratch—something we’ve solved hundreds of times. We should automate those inefficiencies. My passion lies in making sure that no one has to work that inefficiently.

But human insight is essential for creating beauty and meaning in buildings. If computers alone were designing projects, I’m not sure how much beauty would be reflected in them.

Chopson: AI excels at gathering and processing information, but it still needs human intent to create something beautiful. Our goal is to let architects focus more on design and less on coordination. AI enables that shift by handling complexity in ways that humans alone can’t. I think that AI should enable a renaissance in our profession around being able to focus on design more.

And you’ve been experimenting a lot with how AI can help do your job better and quicker. A year ago, you launched Vitras.ai, an AI-powered tool designed to enhance architecture workflows. How does it work? Could you walk us through its capabilities?

Chopson: Traditionally, you buy software to deliver architectural services. We’ve flipped that—our software powers the services we deliver. Vitras.ai can analyze zoning codes, building and fire regulations and will suggest viable solutions.

Then we’ve built a BIM platform, which is like an agentic AI for architecture. You can say ‘Design a floor 10 by 10 meters at this elevation’ and the system understands the request, retrieves relevant code and cost data and produces a structured, traceable output. That’s the brain behind our architecture services.

How did Cove Architecture come about? What made it possible?

Ahuja: This has been a decade in the making. Our first company showed us the inefficiencies in architecture—60-hour weeks shouldn’t be the norm. I’m a mother now, and work-life balance is critical. That drove us to automate the low-hanging fruit over the last five years using our proprietary simulation engines.

As artificial intelligence evolved, we integrated our simulations with AI tools and launched the full-service architecture practice through cove. We partnered with forward-thinking clients like RPL Management, which believed in our AI-powered, but also human-centered approach. Now that we’ve delivered a few successful projects, it’s more about just continuing to help raise the envelope on what’s considered normal in our space.

---

READ ALSO: Using AI to Enhance the Renter Journey

---

Chopson: A year ago, I would’ve said an AI architecture firm wasn’t possible. But recent advances in AI’s ability to understand intent and deliver structured, verifiable outputs have changed that. Now, we can build tools that trace assumptions, reference real codes and assemblies and deliver accurate 3D models—something that wasn’t feasible just a few years ago.

Ahuja: So for example, all the different codes and assumptions about wall assemblies or envelopes or floors for a variety of building types—we have access to their properties. So at any given point we have a really good idea of what the assembly is going to be, what the cost is going to be, what the carbon is going to be. AI paired with our simulation engine paired with our databases are able to do much, much cooler and precise things.

Chopson: And also paired with people. You need experienced individuals who understand the outputs. You don’t want an intern pressing a button and making a building. That would be very bad.

Can you share more about your Atlanta West End project and how AI helped?

Chopson: The original design for that site was eight units, which didn’t make financial sense. Using our AI tools, we redesigned it to support 17 units.

The West End is tricky—it’s bordered by commercial zones and historic districts, with overlapping zoning overlays, parking requirements and height transitions. Normally, analyzing all those constraints would take weeks. But our AI instantly processed the zoning and building codes and iterated designs. We even compared results with a senior architect’s hand calculations, and the AI matched his insights. That enabled us to quickly identify the best solution for both yield and aesthetics.

Can your tool assist with affordable housing development and tax credit compliance?

Ahuja: Absolutely. We use a multi-agent approach to tackle various angles—market studies, zoning, cost, and sustainability. In terms of affordable housing ROI, we can quickly assess what’s feasible based on salary data, land cost and design constraints.

We’ve built agents that optimize for unit count, cost and compliance, while ensuring operating expenses are low through sustainable design. Speed matters—developers often need to justify a project quickly to secure financing or apply for credits. We help make that process faster and more accurate.

Chopson: One example: A developer wanted to build $50 million (worth) garden-style apartments in Kentucky but the return was only between 8.5 and 9 percent. He asked us to help design something that would get him to a 15 percent return. If we can do that, then it gets built. Using our tools, we could evaluate many options quickly to find a viable configuration—something that might take months manually.

So would you say that more developers are embracing AI?

Ahuja: Based on everything that we’ve seen, it’s less about the tools at hand and more about what we as a team are able to provide for them. It’s about better design faster and at competitive prices. I will highlight that they’re most excited about the quality of design. It’s because the quality allows us to add in insights like Patrick mentioned. If we can squeeze in two extra units, the math changes: We can optimize to make sure that the performer gets to a higher degree of return.

Chopson: They don’t care how we do it—just that we deliver on time and with quality. There’s a generational shift happening as experienced professionals retire and delays in permitting are common. We offer the certainty they need.

---

READ ALSO: Resilient Design—4 Apartment Case Studies

---

What are the biggest misconceptions about AI in architecture?

Ahuja: The most common myth is that AI just generates ugly designs at the push of a button. That’s the only exposure many people have—bad generative design.

Chopson: More options aren’t the goal. We’ve moved beyond generative design to what we call collaborative design. Our AI works like a colleague—it helps explore possibilities but within the parameters we define as designers. It’s not about replacing creativity, it’s about enhancing it.

AI is very good at predicting a solution based on a certain set of inputs. And that’s where it’s more like a person than it is. I think pushing a button is absolutely the worst idea. What you want is to say: ‘Here’s what I’m trying to do. What are some things I’m not thinking of?’ for example—it’s a much more fun way to interact with the computer, because that’s how you interact with people.

Looking ahead, how do you expect AI to transform architecture, engineering and construction over the next decade?

Chopson: I think there’s going to be a time, maybe in the next five years or less, when many buildings will be put together using AI as a standard course of business. Right now, I think a lot of professionals are not ready for that, but it’s going to get there.

I don’t think people are going to lose their jobs, but people will be going home on time.