How OpenSpace AI Is Bringing Visual Intelligence to the Construction Industry

Jeevan Kalanithi didn't arrive at construction technology by accident. The CEO and co-founder of OpenSpace AI traces his path back to the MIT Media Lab, where he and his future co-founders Mike Fleischmann and Philip DeCamp studied artificial intelligence, computer vision, and machine learning. After graduating, each went on to found their own companies: Kalanithi launched a consumer hardware company that was acquired by drone and software firm 3D Robotics (later acquired by Esri), while Fleischmann built Bluefin Lab, a computer vision and media analytics company acquired by Twitter.

When the three reconnected, they brought together a rare combination of expertise: a deep understanding of AI and computer vision, experience commercializing technology, and firsthand knowledge of how the construction industry actually operates in the field.

"We didn't solve some neato technical gizmo problem," Kalanithi said. "We solved something that actually moves the needle for field teams."

At its core, OpenSpace is a visual intelligence platform for the built environment. The company makes it easy to capture complete visual records of construction sites using 360-degree cameras, drones, smartphones, and laser scanners.

"If it's happening out there in the field, it's in OpenSpace," Kalanithi said.

The platform transforms raw visual data into actionable workflows, integrating with construction management tools like Procore and Autodesk. Its progress tracking layer converts visual data into metrics like percent complete, schedule tracking, and productivity insights, replacing what has historically been manual, opinion-based reporting with automated, fact-based intelligence.

One standout feature is its AI AutoLocation, which functions like indoor GPS. Since GPS doesn't work inside buildings, OpenSpace uses its 360-degree data to automatically pin observations to the correct location on a drawing, directly from a smartphone. Another is the platform's speed: captured walkthroughs are processed in an average of 15 minutes, giving teams near-real-time access to site conditions.

When asked what makes OpenSpace difficult to replicate, Kalanithi pointed to two things: an obsessive focus on ease of use, and a powerful data flywheel.

"We're psychotically focused on ease of use," he said. "All the promises in the world are useless if they don't ground out better experiences for the men and women actually doing the work."

That focus has driven significant adoption, resulting in over 69 billion square feet of images on the platform and roughly 30 million images uploaded each week. That scale matters because OpenSpace's AI systems are trained on this data, and the more usage the platform gets, the more accurate and capable it becomes, which in turn drives further adoption.

"You can't just write software or replicate it," Kalanithi said. "You'd have to have all this data to get it to work as well as we can."

OpenSpace's customers span the full construction ecosystem: general contractors, specialty trades, and project owners. The platform is used on everything from massive data center builds, to small retail fit-outs and tenant improvements as compact as a 300-square-foot coffee shop.

The company also serves customers in process industries and oil and gas, wherever field execution defines financial outcomes.

"Really, it's for companies who exist in the real world economy, where the physical world is how they make money," Kalanithi said.

Why Construction Is Now a Proving Ground for AI

For years, construction has been characterized as slow to adopt technology. Kalanithi pushes back on that narrative.

"It's not that construction is intrinsically slow at trying technology," he said. "The real reason is that technology has not been built for builders."

In his opinion, traditional software was designed for office workers and people interacting with computers via keyboards. The information that drives construction, however, exists in physical reality, not in documents or spreadsheets. Computers simply couldn't understand that reality until recently.

"Spatial AI and computer vision are the real true unlock," Kalanithi said. "Now we can actually create technology that lives in their world, doesn't disrupt them, and simplifies how their work actually gets executed."

At OpenSpace, that vision is framed around a simple idea: for AI agents to impact the built environment, they need eyes. "We provide those eyes," Kalanithi said. "We allow these agents to look and see what's actually going on."

Visual Intelligence as Infrastructure

Kalanithi sees visual intelligence not as a nice-to-have, but as a new foundational layer in how projects of all sizes get built and managed.

He described this in a telling anecdote: a team member was reviewing reports on a hotel under construction in Mexico. The report indicated 90% completion. When someone was sent to the site to verify, it was 5% complete.

"That is not uncommon," Kalanithi said. "This visual way of working just makes that go away."

Visual intelligence, in his framing, sits alongside ERPs and project management systems as a third pillar of construction software, one anchored to reality in the field and feeding accurate data up through the rest of the organization.

Kalanithi closed with a broader thesis: the construction industry has long operated under conditions of information scarcity and mistrust, leading to inflated insurance rates, fragmented risk, and delayed payments throughout the supply chain.

"I really think there's an industry transformation that can happen where you could have companies that say, 'I'm just gonna buy all the risk because I have the technology and the transparency to execute this work,'" he said. "The trades will get paid on time. They won't have cash flow issues. They can invest in training and people."

It's a sweeping vision for a technology company, but Kalanithi is direct about the stakes.

"If you can move the needle on reducing the cost of managing the real world economy," he said, "that's a pretty big unlock for civilization."