Today we're announcing $60 million in new funding to scale Matic. We didn't raise this on a vision of what might be possible someday. It is an outcome of putting real robots in real homes and solving real problems.
Matic is a product company. We use robotics and AI to solve real customer problems, and we are on a mission to eliminate the drudgery of home chores and take our precious time back.
We've shipped over 6,000 robots. They’ve cleaned over 110M square feet, traveled over 80K miles inside homes, and saved over 31 years of human time. This is the story of how we got here, what we've learned, and where we're going next.
The Question Everyone's Asking
Since the launch of the first disc robot vacuum over 24 years ago, everyone has asked the same question: What comes after Roomba?
Silicon Valley's answer has been loud and clear: humanoids.
Billions are pouring into robots that walk, wave, dance, and do backflips. The demos are stunning. The promises are impressive. The vision is compelling…a general-purpose robot that can do anything a human can do.
We respect that vision. We really do. But we want to lead our customers there too because they keep wondering how will Silicon Valley deliver this future, if we can’t even ship robots that don’t need to bump into our homes, that don’t chew wires, that don’t step into dog poop or cat vomit, or get tangled up in rug tassels? If robot vacuums cannot deal with those pesky adapters on the floor, how would humanoids deal with them?
So we set out to earn customers’ trust starting with a floor cleaning robot that just works.
Eight Years Ago, We Asked a Different Question
In 2017, we didn't ask "What's the most impressive robot we can build?"
We asked: "What's the most useful robot we can ship?"
Turns out those are very different questions.
Mehul and I both come from Google Nest. Mehul led Nest Cam products, and I led computer vision and helped spec out Google Coral TPU. We'd learned the hard way what happens when you prioritize demos and research over products, when you chase headlines instead of solving customer problems.
So we started with a simple observation: People hate vacuuming and mopping. Roomba tried to solve it, but customers remained frustrated and disappointed. They got stuck. They hated "pre-cleaning" before disc bots could clean, babysitting them before disc bots made a bigger mess, and constantly rescuing them when they got stuck.
So we asked, what if we could build a robot that actually worked? One that you could trust to handle your home as it is, without needing to pre-clean or babysit?
We thought it would take three years and we'd ship in 2020.
Well... we were wrong. Very, very wrong!
Real Homes Are Chaotic, Messy, and Cluttered
Here's what we learned in those first three years: Lab environments lie to you.
In a lab, you can control everything. The lighting is perfect. The floors are clear. Obstacles are predictable, pre-arranged, pre-defined. In a lab, you don't have your dog chasing your kiddos or kids leaving shoes, toys, food crumbs, and umpteen other things anywhere and everywhere.
You have that in real homes. Legos everywhere. Dog bowls in random places. Furniture that gets rearranged. Infinite types of floors, wires, rugs, carpets, tassels, colors, stairs, wall-to-wall glass, mirrors, and doors. Skylights, sunlight through the window, night lights, recessed lights, yellow lights, white lights, shadows, and the list goes on.
Every home is different, unique. Every floor is different, unique. The variables are endless, unique.
The result is a policy that works at 95% accuracy in labs drops to 60% accuracy in homes.
So we did something different. We started putting bots in our own homes and began shipping to real customers early on without the hype or overpromise. We assembled robots by hand in our Mountain View facility. We sent them to homes with very patient, willing customers. Some customers got fed up, got mad, and returned Matic. But we listened to all the feedback. We iterated. We shipped again and again.
It was slow. It was unglamorous. It was tedious. While competitors were announcing billion-dollar valuations and flashy partnerships, we were holed up in our tiny office, hand-assembling robots and testing them in houses with messy kids and furry pets.
But Matic got better. Performance improved. Customers started depending on them, and some even refused to send beta robots back even when we had new ones for them.
We uncovered "earned insights."
Digital AI vs. Physical AI: The Autonomy Gap
We learned that in digital AI, collaboration is fine. If a coding assistant gets something 90% right, that's massively helpful. You work together to finish the last 10%.
But physical AI is different. Physical AI is about delegation, not collaboration.
A robot taxi that gets you 99% of the way home isn't helpful. You're still stranded. A robot vacuum that cleans 95% of your floor and gets stuck on a sock isn't autonomous. You still have to babysit it.
The bar for success in physical AI isn't just higher. It's exponentially harder.
Real autonomy isn't about being "collaborative." It's about being invisible. Like Alfred in Batman, it should just take care of our needs without us needing to ask.
That's the bar we're shooting for. You tell the robot to clean your home once, and it just does it every day, just the way you like it, without you needing to tell it ever again.
What We Built
That’s the goal for Matic.
That’s why it sees your home with computer vision and AI the way you do. It understands your space as it actually is, with all its idiosyncrasies.
It knows to vacuum hard surfaces and rugs, but only mop hard surfaces. You can tell it to remove dry stains. It doesn’t get stuck, doesn’t require pre-cleaning or babysitting. It's quiet enough for nap time (ask our daughters). And doesn’t scare the bejesus out of our doggies.
And it's built in America. Every single robot is assembled, built, tested, and quality-checked in our Mountain View facility.
WIRED gave it 10/10. Tobi Lütke called it "incredibly delightful." Many customers have called it the best purchase since the iPhone.
The Customer Vote
Here's something that doesn't show up in press releases but shows up as a bottom line: customer love.
We get messages like this all the time:
"I bought Matic for my wife for Mother's Day and caught all kinds of flack for getting a robot vacuum as a gift. Now she absolutely loves it and raves about it to her friends."
Handwritten notes of customer love and appreciation
These stories don't go viral on Twitter. They don't generate headlines.
But they're everything. Because here's the truth: Customers don't buy humanoids or robotic foundation models. They buy products that solve their problems.
They don't care about your transformer architecture or your foundation model or your end-to-end NN resulting in human-like locomotion. They just care if the robot is solving their problem, cleaning their floors…thoroughly. They want to live in a perpetually clean home, with perpetually clean floors. But they don’t want to clean it. They don’t even want to think about it. They simply want something that just works at a relatively affordable price point.
So we focus on what customers want. Our philosophy is very simple: if customers win, we win. That’s it.
Why This Round Matters
We're scaling fast. We have overwhelming customer love and reviews. And we're still only 50% through what we consider our "minimum lovable product" roadmap. And Matic's just step one in our quest to make the Rosie the Robot dream a reality.
This funding lets us:
Scale manufacturing to get Matic into more homes
Keep iterating based on real-world usage and customer feedback
Build the next generation of features (voice control, advanced stain detection, and more)
Start working on the next set of products
But more than that, it validates something we've believed from day one: Insanely delightful products that customers love are the only way to build iconic companies.
Made in America (And Why It Matters)
We're often asked why we manufacture in the US when it would be cheaper to go to Asia.
The answer is simple: Building in America allows us to iterate faster, scale faster, and innovate faster.
In addition, we focused on privacy and security of users’ data. We never understood why we have to jeopardize privacy and security of home, our children, and our families just to keep floors clean.
That's exactly what we're doing while helping reindustrialize America.
The Philosophy
Eight years in, here's what we believe:
Customers vote with their wallets, not with hype. Products win by solving problems and delivering jaw-dropping delight, not by looking futuristic.
Real autonomy is invisible. If you're still babysitting it, it's not “fully” autonomous yet.
Physical AI is about delegation, not collaboration. 99% isn't good enough.
Must iterate in real homes. You can't learn what matters in a lab.
American innovation is alive and well. We just need to build things people actually want.
So What Comes After Roomba?
It's Matic. We're incredibly proud of what we've built. But we're even more excited about what's coming, and we can’t wait for our customers to experience it.
Want to experience it for yourself? Matic is available at maticrobots.com. Every robot is designed and assembled in Mountain View, California.
— Navneet
P.S. We're hiring. If you want to help us build delightful and iconic home robots that solve real problems for real customers, we'd love to hear from you.
→ See open roles: maticrobots.com/careers → Don't see your fit? Email apply@maticrobots.com with your resume and what you're looking for.
P.P.S In the Imitation Game movie, the great Alan Turing says, "Sometimes it's the people no one imagines anything of, are the ones who do the things that no one can imagine." Turns out it's true for Robots too: sometimes it’s the robots that no one imagines anything of, are the ones that bring about the delightful future that no one can imagine.
Vacuuming
Vacuuming