Who Is Steve Ruiz?
Steve Ruiz started as a fine artist — pens, paint, drawing, painting. He got into technology through design, then design tools, then the deep problem of how to build a canvas that actually works on the web. Four years ago, he gave himself a couple of months before taking a job at Adobe. Instead, he finished an open source project, it got popular, it kept getting popular, and now he runs tldraw — a 15-person startup in London that builds the canvas infrastructure powering Shopify, Google, ClickUp, Autodesk, and Observable.
In 2024, he built Make Real — one of the first tools to turn hand-drawn sketches into working code using GPT-4 Vision. It went viral. It predated Lovable and Bolt. When his investors asked why he didn't just build the obvious app, his response became one of the most quoted lines from his interview: "Yeah, you sound like my investors."
He chose infrastructure over product. A thousand applications instead of one. And he's still having more fun than almost any founder we've talked to.
The Archetype: The Creator
The Creator
The Jester
The Reward
Steve's primary archetype is The Creator — someone driven by the act of making, not the outcome of selling. This isn't a founder who talks in revenue milestones, competitive moats, or market positioning. He talks about geometry systems with the pride of a woodworker describing a dovetail joint. He tests everything against whether his seven-year-old daughter can play with it. He calls his experimental projects "toys" and "playgrounds" even when they have real commercial potential.
His secondary archetype is The Jester. The playfulness isn't decoration — it's structural. He turned AI agents into fairies with customizable hats and leg lengths. His Make Real system prompt described the AI as "a senior developer, you're 4,000 years old." When asked about AI risks, he replied: "I hope so. Or else, paperclips." He uses humor to keep things light while saying serious things, and it's a big part of why his interview is one of the most watchable we've produced.
"The creativity balances all the other challenges and stuff like that."
The Hero Match
Hephaestus — Greek God of the Forge
Steve is Hephaestus — the divine craftsman who built the tools the other gods used. Hephaestus didn't rule Olympus. He didn't lead armies. He built Achilles' armor, Hermes' winged sandals, and the automata that guarded his workshop. His power was in the making.
tldraw operates the same way. Shopify builds internal collaboration tools on Steve's canvas. Observable uses it for data visualization notebooks. ClickUp uses it for whiteboarding. Google uses it for internal tools. They all build on his infrastructure the way the gods wore Hephaestus' creations. And like Hephaestus — the only Olympian who worked with his hands — Steve's fine art background means he literally started with craft before he ever wrote code.
The deepest part of the parallel: Hephaestus was genuinely unbothered by the gods who got the glory. Steve doesn't want to be Lovable. He wants to build the thing Lovable runs on.
Hiccup from How to Train Your Dragon — DreamWorks, first film
Steve is Hiccup before Hiccup becomes chief — the kid in the blacksmith's workshop building things nobody asked for. Prosthetic tail fins. Automatic bola launchers. Riding saddles that shouldn't work but do. Hiccup doesn't fit the Viking mold (warrior, conqueror), and Steve doesn't fit the typical founder mold (scale fast, capture market, become Lovable).
The daughter connection makes this uncanny. Hiccup's relationship with Toothless has the same playful-craftsman energy as Steve testing fairies, tldraw Computer, and every new feature against "can my daughter play with it?" Innovation through observation and play, not ambition and strategy.
"I'm building for it one way or another."
The Story Behind tldraw
The Founder's Journey ↔ The Company's Journey
Fine artist → design tools → gave himself "a couple of months" before Adobe → open source project got popular → "kept getting popular" → turned down the Lovable path → built the canvas infrastructure layer → now exploring what happens when AI agents live on a canvas.
Open source canvas library → SDK for whiteboard apps → Make Real goes viral → infrastructure powering Shopify, Google, ClickUp, Observable → fairies (AI agents on canvas) → tldraw Computer (executable visual workflows) → exploring canvas as the default UI for AI collaboration.
The same archetype drives both: The Creator who can't stop building rebuilt not just the product but the entire canvas category. Steve's personal evolution from artist to infrastructure founder mirrors tldraw's evolution from a drawing tool to the engine underneath a thousand products. The bet was always the same — make something flexible enough that others can surprise you with what they create on it.
How Steve Leads
Steve owns his calls with unusual clarity. When he narrates past decisions, it's always "my bet was" — singular, personal, no hedging. Not becoming Lovable, keeping tldraw as infrastructure, charging $25 for fairies as a marketing experiment — these are decisions he stands behind without apology.
But he's genuinely humble about what he doesn't know. He admitted fairies weren't profitable ("if you ignore all labor, then no"). He acknowledged that Make Real's API key requirement locked out 80% of visitors. He said "I don't want to contribute to the bad AI generated content out there." This is a conviction-driven founder who doesn't need to be right about everything — just about the one big bet.
The core tension: Craftsman vs. Businessman — Steve's energy comes from the collision between wanting to build beautiful, playful, deeply-crafted things and running a company that needs to sell licenses and please investors. He resolves it by making the craft the business strategy — "marketing by building." The tension is productive. It's what gives tldraw its character.
Founder Superpowers
Translating Complexity Into Play
Steve takes concepts that should be intimidating — AI agents coordinating in parallel, WebGL shader pipelines, canvas geometry systems — and makes them feel like toys. AI agents became fairies with customizable hats. A visual workflow tool became "probably the world's worst programming language." He got the H2O molecule wrong on camera during a chemistry tutor demo and laughed about it. This isn't just communication style — it's design philosophy. He builds things that are technically deep but experientially light.
Seeing Patterns Before Products
When Make Real went viral, every signal said "build the app." Steve saw something else: "This didn't feel like a product. It felt like a pattern — a thousand applications rather than just one." That pattern-recognition is his rarest capability — resisting the obvious product because he can see the platform underneath it. Grant Cott's liquid simulation, Observable's data notebooks, the Chinese bible study community: each validated the abstraction by building something Steve never designed for.
Marketing by Building
Steve doesn't do traditional go-to-market. Make Real was his marketing. Fairies was his marketing. tldraw Computer was his marketing. Each experimental project doubles as a proof of concept that attracts exactly the right developers. He charged $25 for fairies not for revenue but to see "to whom this would matter." The demos ARE the distribution — and they only work because each one is genuinely impressive.
What It's Like to Work With Steve
Based on an hour-long unscripted conversation, a few things stand out about what Steve would be like to work with day-to-day.
He's a show-first thinker. His instinct is to reach for a live demo before an explanation. "Let me show you something" came up repeatedly — and each time, the demo communicated more than words could have. If you work at tldraw, you're probably prototyping ideas, not writing memos about them.
He's playful with serious work. AI agents became fairies with customizable hats. A chemistry tutoring demo involved him getting the H2O molecule wrong and laughing on camera. The system prompt for Make Real described a 4,000-year-old developer. If you thrive in environments where craft and humor coexist, this is the team.
He trusts his team with execution. Steve owns the strategic bets ("my bet with tldraw was"), but the building is always "we." He credits specific people — the Figma designer who first prototyped Make Real, Observable's founders who chose tldraw for data visualization. He's not a micromanager. He's someone who makes the big call and then lets the team run.
"I have a blast using this app."
Why This Matters (For You)
If You're Building an Application That Needs Canvas or Whiteboard Features
Steve built tldraw because he saw the web's canvas problem — traditional graphics rendering doesn't play well with web technologies, collaboration is hard, and most developers end up fighting the framework instead of using it. tldraw solved this by building a canvas infrastructure SDK (not a finished app) on HTML/CSS primitives, making it accessible to developers building inside Shopify, ClickUp, Google, Autodesk, and Observable. If you're a product team or developer evaluating canvas infrastructure for your application — whether it's collaborative whiteboarding, visual design, data visualization, or spatial interfaces for AI agents — tldraw's philosophy is worth considering: choose infrastructure that gets out of your way and lets you layer your own domain expertise (like Grant Cott layering WebGL physics on top of tldraw's geometry system). Steve's bet was that the best canvas would be the one built for builders, not users, and the customer list backs that up.
If You're an Engineer Building AI-Powered Tools
Steve's infrastructure-over-product decision is worth studying. When Make Real went viral, the obvious move was to productize it. Instead, he recognized the pattern underneath: any application that takes visual input and feeds it to an AI model is doing the same thing. By building the canvas layer, he positioned tldraw as the foundation for Lovable, Bolt, and every AI design tool that followed — without competing with any of them. The engineering lesson: when you see a pattern that could be a thousand products, build the platform, not the app. And test your abstraction by seeing if strangers can build things you never imagined (Grant Cott's liquid simulation was Steve's proof).
If You're Early in Your Career
Steve almost took a job at Adobe. He gave himself "a couple of months" to finish an open source project first. That project became tldraw. His book recommendation — Red Mars by Kim Stanley Robinson — is about people who live 500 years and develop "a complete fearlessness about learning something new." A physicist spending 12 years studying biology because they have time. Steve sees AI tools as giving everyone that same fearlessness: "I never thought I would be doing firmware for little development boards, but I'm doing that now." The career lesson from Steve isn't "follow your passion." It's: give yourself room to finish the thing you're working on, and don't be afraid to learn something that seems unrelated.
If You're Considering Joining tldraw
Steve leads with craft, not metrics. He tests products with his daughter. He calls experimental projects "toys" and doesn't oversell them. He credits team members by name. He describes a 15-person office-first team in London that works on problems at the intersection of creativity, infrastructure, and AI. If you want to work somewhere that values deep engineering and playful experimentation over growth hacking and pitch-deck metrics, Steve's interview gives you a clear picture of the culture. The company is actively hiring — check tldraw.dev for current roles.
Join tldraw
Now that you know how Steve Ruiz leads, see if there's a role for you.