If you aren’t a current user of Matterport’s hardware or software, you might know them mostly from the virtual tours common in real estate, which are helpful for brokers, homeowners, and buyers to explore properties from afar before scheduling a showing. But in recent years, Matterport has broadened that scope, moving beyond real estate and finding a fit in niches left open by other imagery providers.
By positioning themselves as user-friendly reality capture hardware and software company with a low barrier to entry, Matterport has unlocked use cases where simply getting inside, capturing a scan, and producing a shareable visualization is what matters most. In applications where a “close look” is more valuable than measuring down to the millimeter, Matterport has room to shine. I was able to catch up with Stephanie Lin, VP and GM of Matterport’s Operate Solution to learn more about what’s changed and where the company is heading.
Commercial and Facilities Applications
Imagine you are a developer considering purchasing a warehouse, factory, or other industrial space. How can you get a sense of whether the facility will be a good fit when it is currently filled with equipment or difficult to photograph? Traditional 3D scanning methods can be time-consuming, often capturing more detail than you need, and the clutter can obscure the actual layout.
This is where Matterport’s AI-powered Defurnish feature comes in. With a single click, it removes anything “not bolted down” from a scan—furniture, machinery, even pipes—leaving behind a clean, rendered version of the space. The result is a digital twin that is easy to dimension, convert into a floorplan, and quickly analyze.
Technically, the feature works by applying semantic segmentation to identify which elements are furnishings or obstructions. Once identified, they’re erased, and the software fills in the missing geometry and textures using image inpainting and mesh repair. It’s part of Matterport’s broader Project Genesis initiative, which aims to apply generative AI to not just “defurnishing” but also redesigning spaces, virtually furnishing them, and even reimagining their style.
AEC and Design Applications
Architects, engineers, and contractors often need rapid as-built documentation to inform a remodel or retrofit. Matterport provides a fast-track option that balances speed and accuracy. Instead of coordinating a detailed lidar survey, teams can capture a space with Matterport and extract floor plans, 3D models, and visuals that are “good enough” for early design decisions.
The ability to toggle between furnished and unfurnished states helps teams quickly communicate possibilities with clients, while integrations into CAD and BIM software ensure the data isn’t locked in Matterport’s ecosystem. For design professionals, this lowers friction in early-stage planning and helps clients visualize outcomes.
Another AI-driven tool is the ‘merge’ feature, allowing users to combine multiple 3D scans into a single, seamless model. This is especially useful for large properties or projects that require scanning in phases, since it eliminates the need to capture everything in one session. By merging models, teams can build a more complete digital twin, simplify navigation, and deliver a unified output for stakeholders without sacrificing detail or accuracy.
A Broader Vision: From Tours to Intelligence
Matterport’s AI push is not just about removing furniture. Features like auto-generated property descriptions, customizable floor plans, and CAD-ready exports are aimed at turning a simple scan into a package of insights and deliverables. The company envisions a workflow where the same digital twin can support marketing, design, operations, and long-term management—bridging industries from real estate to manufacturing.
While challenges remain (AI inpainting still risks errors, and some industries require higher precision than Matterport offers), it is clear that the company is carving out a middle ground between consumer-friendly imagery and industrial-grade scanning, making “good enough” data far more useful with the help of AI.
