Ep 55: Social Media & AI for The Interior Design Community
Social Media & AI for The Interior Design Community
Canvas I.O. CEO Sarah Bird on 17 years of LiDAR technology, scanning a room in 10 minutes instead of 2 hours, and why AI can replicate patterns but will never replicate taste.
- Canvas I.O. turns an iPhone Pro or iPad Pro into a field measurement tool. Scan a room in minutes and receive a detailed, design-ready CAD file native to Chief Architect, SketchUp, Revit, or 2020 Design — typically within a quarter to half inch of accuracy for single rooms.
- The software is a hybrid model: AI extracts walls, planes, and geometry; humans add the critical details. Outlets, light switches, baseboards, and trim are reviewed and added by the team before the file ships. Accuracy is not sacrificed for speed.
- Canvas is the real-world connection that generative AI lacks. Every AI visualization tool generates beautiful imagery — but none of them know what the actual room measures. Canvas solves that foundational problem.
- Showing up with Canvas at a first client meeting is a positioning statement. Scanning the space in 10 minutes while the client watches is a visible, concrete demonstration that this is a tech-forward firm — before a single design decision has been made.
- AI can replicate patterns. It cannot replicate taste. The distinction matters for designers thinking about where their irreplaceable value lives: in judgment, material knowledge, client communication, and the ability to sense what a space needs.
Sarah Bird is the CEO of Canvas I.O., the LiDAR-powered scanning and CAD modeling platform built for interior designers, architects, design-builders, and remodelers. A former COO, President, and CEO of a marketing technology company she scaled and sold, Sarah came to Canvas after a sabbatical and was drawn to its unique intersection of science, art, and deep technical complexity. Canvas holds over 20 patents developed over more than a decade.
From Kickstarter Hardware to Interior Design Essential
Canvas I.O. has a longer history than most designers realize. The foundational technology — mobile LiDAR sensors — traces back to 2008, when the original founders built the first company to ever create mobile LiDAR hardware. Long before Apple put LiDAR in the iPhone Pro, Canvas's predecessors were doing a Kickstarter for a tiny LiDAR unit that plugged into an iPad. The company has held more than 20 patents through more than a decade of development.
What we know as Canvas today became its own focus around 2022, when it became clear that the real problem to solve was not hardware — it was the labor of field measurement and modeling. Measuring a space by hand, entering those measurements into design software, tagging every element correctly, and producing a usable as-built model is genuinely painful work. Canvas is built to eliminate that pain.
"I just want it to feel like a relief. I want designers to think: thank God Canvas is here. I don't have to worry about this."
— Sarah Bird, CEO of Canvas I.O.Sarah came to Canvas after a career building and selling a marketing technology company. She was drawn to it specifically because it operates at the intersection of science and art — and because the customers are people who understand that the spaces they design change how human beings feel, relate to each other, and move through their lives. That is not a small thing to be in service of.
What Canvas I.O. Actually Does
At its core, Canvas I.O. uses the LiDAR sensor built into the iPhone Pro and iPad Pro to scan a physical space and convert it into a design-ready CAD model. The process has two tiers — a fast instant draft for conceptual work, and a detailed as-built file for serious design and construction use.
When to still use traditional measurements: Canvas is accurate enough for nearly all design work — but Sarah recommends field-verifying critical dimensions for windows, cabinets, and countertops. An inch of variation in a floor plan is fine; an inch of variation in a cabinet cutout is not.
The detailed as-built files are native to the major design platforms:
The goal is native files for every major platform — so that the objects in your Canvas file match the object types your software expects, rather than requiring cleanup after import. The instant draft exports as a USDZ file for quick viewing without design software.
How to Get Started — The Full Workflow
Canvas is free to download and free to practice with. Two projects per month can be scanned at no cost. There is no subscription required to start — you only pay when you order a detailed as-built file, and billing happens when the file is delivered, not when you upload the scan.
Canvas as the Real-World Connection for Generative AI
Generative AI tools for interior design are extraordinarily capable at visualization — take a description or a photo, get back an image of how a space could look. The limitation is that they are generating onto nothing. They do not know what the room actually measures, where the HVAC vent is, how wide the doorway really is, or which walls are load-bearing.
Sarah's vision for Canvas is to serve as that foundational layer — the real-world input that every visualization tool currently lacks. Instead of starting an AI visualization from scratch, start from a Canvas model of the actual space, and build your design work on top of accurate physical reality. Every furniture suggestion, every layout option, every rendering is constrained by the actual dimensions of the room rather than by assumptions.
"I want to be the part of the process where we make the real world designable — without compromising accuracy or detail — so that whatever visualizing tool you use, you start from this is how the room is today."
— Sarah BirdThis also has a practical business development application. Showing up to a first client meeting, scanning the space in ten minutes while the client watches, and sharing a 3D model of their home before leaving — that is a visible, concrete demonstration of a tech-forward firm. Clients do not need to understand what LiDAR is to be impressed by watching it work in their own living room.
What AI Cannot Do — And Why Designers Should Know the Difference
Sarah is genuinely enthusiastic about AI — and also clear-eyed about where its capabilities actually end. The distinction she draws is between pattern replication and taste. Generative AI learns from examples and can produce outputs that match those patterns. That is genuinely useful. It is not taste.
Taste requires understanding why something works, not just that it looks like other things that have worked. It requires knowing how light changes the perceived color of a material at different times of day in a specific room with specific window orientation. It requires feeling the weight and texture of a fabric and knowing it will read as too heavy in a space that needs to feel airy. None of that is available to a model that has only ever processed images.
"AI doesn't have taste. It has copycatting. It's learned the pattern and it can replicate the pattern — which is not the same thing as taste."
— Sarah BirdJenna added the framing she uses consistently across the podcast: designers are the trendsetters. If something has not been made yet, AI does not know it exists. The people generating the new visual language of design — the designers doing work that has not been done before — are, by definition, doing something AI cannot replicate because it has not seen it yet.
Sarah also raised the possibility that AI could modestly elevate the baseline quality of design for people who cannot currently access professional services — basic feedback, layout suggestions, proportion checks. She sees that as a complement to professional design, not competition for it. The clients who want partnership, judgment, material knowledge, and a human who can sit in a room and sense what it needs — that tier of the market does not disappear when AI can suggest lamp placement.
Disclaimer: This blog was written using AI as a recap from the recording then edited by the author for accuracy and details.
