The other day, my 8-year-old son looked up from his giant pile of Legos* and asked, “What exactly do you do at work all day, besides talk on the phone?”
Fair Enough.
Well, I could have told him that my company builds exabyte-scale, software-defined, cloud-native file data platforms for the reasoning revolution. But I knew his eyes would instantly glaze over. And let’s be honest, sometimes we get that reaction from adults, too.
True mastery of a subject means being able to turn complex ideas into something simple and clear. As Albert Einstein famously said, “If you can’t explain it simply, you don’t understand it well enough.”
Instead, I looked at his messy playroom and realized that the companies we help have the exact same problem he does: they have way too much “stuff” (data), they want to play with it as fast as possible (compute), and they don’t want to clean up or organize it.
Here is how I explained the three biggest problems my company, Qumulo, solves using the most important thing in his world: Legos.
*Important Note: Ok, let’s get this out of the way early. Yes, I know the plural of Lego is still Lego, and not Legos (it’s a mass noun). I am also aware that it should be spelled in all capital letters. I just can’t bring myself to it;
1. The Ultimate Elastic Play Table & Bottomless Toy Chest
(Solving Cloud Costs, “Noisy Neighbors,” and Sharing)
I pointed to his room. “Imagine you have a giant, bottomless toy chest where you keep absolutely everything you own. But when you actually want to play with your toys, they appear on your table, organized and ready to play with at the snap of a finger.
Now, imagine you invite five friends over, and everyone wants to build their own spaceships at the exact same time. You suddenly need a much bigger table so everyone can build without delay, making everyone that much faster to build the most impressive spaceship.
With older, traditional toy systems, if you wanted a bigger play table for your friends, you were forced to buy a giant, expensive new toy chest to go with it, even if you didn’t have any new toys to store in it! Worse, when your friends went home, you couldn’t return it. You were stuck permanently keeping that giant table and chest in your room, wasting your allowance on space you only needed for one afternoon.
Now, imagine you suddenly get a lot more toys, like after your birthday; you might want to make your toy chest bigger. Or perhaps you decide to donate some of the toys you have outgrown. With a new and improved toy system, the table and the toy chest grow separately. You can have a huge toy chest with a small table, or a giant table with the same-sized toy chest. Each expands only when needed and shrinks back down when it is not.
“But wait,” I told him, “what if your brother comes in and wants to play at the same time? Usually, you two bump elbows and argue over space. We use magic invisible walls so that you and your brother can play on the exact same expanded table, pulling from the exact same toy chest, but you can’t see or bump into each other.”
The Grown-Up Business Translation
In the business world, older storage systems force companies to buy massive amounts of expensive, permanent storage capacity just to get a little more speed. Because performance and capacity are glued together, companies are forced to overpay for peak demand.
Our technology solves this by decoupling performance (compute) from capacity (storage). It allows businesses to dynamically scale their performance up when thousands of users or AI applications need to access data at once, and then shrink it back down when the rush is over, so they only pay for exactly what they use.
Additionally, the “invisible walls” use cryptographic isolation to enable multiple departments to securely share the same equipment without ever becoming a “noisy neighbor.”
2. The Infinite Playground & the Robot Hunt
(Solving GPU Scarcity and Cloud Cost Sprawl)
“Sometimes,” I continued, “your bedroom play table just isn’t big enough for the massive Lego city you want to build. You need to rent out a giant gymnasium. That giant gym is the public cloud.”
“But renting a giant gym is expensive!” he said.
“Exactly!” I replied. “And the longer you rent this gym, the more money you have to spend. But what if you hired a bunch of robots to help you build faster?”
“This can get a bit tricky,” I told him. “To build a giant city really fast, you need to borrow super-rare, lightning-fast robot builders. These robots are incredibly hard to find. One might pop up at the North Gym on Monday, and another at the South Gym on Tuesday. You have to constantly hunt for them!
With older toy systems, if you finally found a robot at the North Gym, you had to pack up your entire heavy toy box and drive it all the way across town. While you were driving, you had to pay the robot out of your allowance to just sit there and wait for your toys to arrive. Or, to avoid waiting, you had to buy five identical Lego sets and leave them at five different gyms just in case a robot showed up, which costs a fortune!
That’s why we built a magical set of rules for the gym. You only need ONE toy box. When a robot builder becomes available at any gym in the city, our technology lets that robot instantly connect to your toy box and start building immediately, without ever moving the heavy box at all.
You and your brother could even find two different robots at two different gyms and build from the exact same toy box at the exact same time! And if you suddenly need room for a million more Lego bricks, we stretch the gym to be huge, but the moment you are done playing, it shrinks back down to a normal size so you don’t waste your allowance.”
The Grown-Up Business Translation
In the business world, companies use the public cloud to rent super-fast GPUs to train artificial intelligence models. However, GPUs are incredibly scarce, and their availability constantly shifts between different cloud data centers (known as Availability Zones).
As a result, companies are forced to “hunt” for GPUs. They are forced to either pay for expensive GPUs to sit idle while they spend days copying their massive datasets, or pay to store duplicate copies of their data in every single zone, just in case a GPU becomes available.
Our cloud technology solves this “GPU hunting” problem by enabling workloads to run across multiple zones simultaneously without ever duplicating data. The data exists in one central, durable place, but compute from any zone can attach to it instantly.
If GPUs are available in one zone on Monday and a different zone on Tuesday, the company can start working immediately without moving data or rebuilding infrastructure. Furthermore, it allows businesses to aggressively scale performance to meet a deadline, then scale it back when the peak has passed, ensuring they never pay for idle headroom.
3. The Magic Lego Shelf
(Solving Slow Data Movement and Version Conflicts)
“Finally, let’s talk about visiting Grandma’s house,” I said. “Normally, if you are at Grandma’s and want to build with your massive Lego collection that lives back at our house, you have to wait for someone to pack it up and drive it over to you. That’s slow and annoying.”
Instead, we built magic Lego shelves that connect our house directly to Grandma’s house, or anywhere else you want to play. You sit at Grandma’s, reach for a Lego spaceship, and it appears in your hand, even if it was at our house a fraction of a second ago.
And if your brother, who is back at our house, touches your spaceship or changes it in any way, you’ll know! If he swaps a blue wing for a red wing, you will see the red wing. You’ll never accidentally grab the old version.
“Here’s how the magic works. Our house is where the giant pile of real Legos actually lives as the ultimate source of truth. The shelf at Grandma’s house acts as a super-smart assistant, keeping the Legos you use most often right next to you and even learning which bricks you might want next, so it can grab them before you even ask.
And the magic tunnel between them is incredibly fast because instead of shipping the whole heavy Lego bin every time something changes, it only sends the exact small bricks that were swapped or moved.”
The Grown-Up Business Translation
In the grown-up world, this solves the “ghost problem” of outdated files, slow transfers, and working across dispersed locations.
A massive, central data center acts as the single source of truth for all your company’s information. The remote location uses a predictive intelligence engine that learns which data you use most often and pre-fetches it to your local environment, delivering lightning-fast speed anywhere.
Instead of copying and sending entire datasets over the network when changes are made, the system streams only the exact tiny pieces of data that have changed. This means that no matter where teams are in the world, they never wait for large amounts of data to copy, and everyone is always working on the exact same, most up-to-date file at the same time.
Take 3 more minutes to read this other blog: Multi-Region Collaboration Done Right: Breaking Distance Barriers with Rel-Time Data Access. This blog presents real-world proof of a media studio successfully collaborating on a project from two different cities: London and Stockholm. Editors could jump to any frame and see it load almost immediately, with full playback reaching smooth 24 fps within moments.
4. When the Pieces Finally Click Together
(Understanding Means Simplicity)
He thought about it for a second, seemingly unimpressed, nodded, and went back to building his spaceship. Much like explaining technology to adults, sometimes you just hope the message landed. Or maybe, if I’m being honest, I just didn’t explain it simply enough. (It’s tougher than you think)
Later that night at dinner, my wife asked him, “So… what does Dad actually do?”
“Dad builds magic so nobody has to wait.”
She laughed. “Oh, really? What does that mean?”
He shrugged.
“It means my Legos are always safe, always super fast to get to, and always ready when I want to build with them, even at Grandma’s. I don’t have to wait, I don’t lose pieces, and my brother can’t steal the good ones.”
He may not care about exabytes, caching, or cloud architecture yet, but he understands the part that matters most: when technology is done right, everything just works. No waiting, No Complexity, No Fuss, and certainly not having to share with anyone if you don’t want to, especially your brother.