The College Font Collection That Started Everything
Senior year, art school. My friends and I were obsessed with fonts.
We'd trade font libraries on burned CDs and flash drives, swap recommendations, spend hours digging through beautiful (and hideous) typefaces looking for the perfect one for our projects. We had thousands. We all had access to incredible typography, but finding the right font when you sat down to work was a different story.
Scrolling through endless font lists in Photoshop. Trying to remember which designer recommended which typeface. Spending entire afternoons browsing just to end up using Helvetica again.
The problem was right there, quietly making every designer I knew bite their nails as they started a project. But it was how it is. Design is hard.
The Thoughtbot Moment: When It Became Real
It's 2008. Fresh out of college, Designer #2 at the world-famous Thoughtbot software consultancy office in Boston.
We're sitting around a table brainstorming side projects. I was the only designer in a room full of brilliant programmers—Designer #1 lived in New York—which meant I had a unique perspective on how broken font selection really was.
Everyone else built software for a living and understood databases, APIs, coding user interfaces, but none of them had ever spent three hours trying to find the right typeface for a client project.
That's when it started to occur to me: maybe I could build something to solve this.
The Vision Before the Technology Existed
Here's what's wild about 2008 and 2009: Web fonts didn't really exist yet.
@font-face was experimental. Most of us were still exporting images from Photoshop to put custom typography on websites.
And Dropbox was taking off, and it inspired me to think about how many fonts I'd lost moving between computers over the years.
So my first mockups were basically "Dropbox for fonts"—a visual browser that synced your font library across computers and a cloud. I posted skeuomorphic wireframes on Dribbble (because Dribbble was hot and new) — complete with wooden textures inspired by those old lettercase drawers that gave the project its name.
The challenge: I'd taught myself to code using Ruby through Rails, but had no idea how to make a menu bar Mac app.
Luckily, I (vaguely) recall, MacRuby let me build Mac apps in a language I actually understood. I was able to create working prototypes, even if they weren't quite polished enough to launch. Of course, Electron didn't exist until 2013, and I was trying to reverse-engineer Dropbox's syncing algorithms as a self-taught developer who'd been coding for maybe two years.
The prototypes worked, mostly — but I kept hitting walls around the final polish to make it great.
The Long Evolution: Waiting for the World to Catch Up
2009: I built prototypes but kept hitting syncing or app problems I couldn't quite crack. I was a growing A Good Company, too, so client work kept pulling me away from side projects.
2010-2012: I got interested with the organization angle—font "playlists" for projects. Revolutionary idea at the time, but I was still struggling with technical constraints and juggling projects, for sure.
2012-2015: Social networks were exploding, so I pivoted to the social angle. What if you could follow your favorite designers and see what fonts they're using? Celebrity designer culture was wholesome back then—everyone was teaching, sharing, lifting each other up. I built working prototypes with social features, but couldn't quite finish.
2020: I started working on an algorithm to auto-categorize fonts by analyzing their points and structure. Before AI really took off, I built something that could classify fonts with 80-85% accuracy based on their structural geometry. But it hit a wall with unusual fonts—reverse italics, handwriting, display fonts with unconventional structures. It wasn't good enough.
2022: The breakthrough finally came. With the right positioning, large language models got sophisticated enough to understand imagery and design context. I started being able to make prototypes that got me back results that were not just "fonts that look similar" but "fonts that work for your specific project, audience, and goals."
The AI Breakthrough: Finally, the Missing Piece
Somewhere in 2022-ish, learning about building products with AI inherently at their core (somnia, a dream processing app, a shadcn/ui theme generator, etc) I realized modern AI could understand a lot with the right guidance and tools and context — not just "show me sans serifs that look like Helvetica" but "I need a font for a tech startup's landing page targeting enterprise clients."
This opened up a new iteration of prototypes I'm still working on, and so excited about. The challenge is teaching machines about readability, brand alignment, emotional resonance—all the contextual factors human designers intuitively consider.
The League Insight: 16 Years of Typography Education
Now, through all this, I've been running The League for equally as long.
They're very different projects, but running The League taught me something crucial: the hardest part of typography isn't making fonts—it's choosing them. After 144 billion font views and making 20+ courses/workshops, and constant designer questions about font pairing, I realized that even advanced designers often work instinctually — without formal process.
This insight directly shaped how I'm approaching Lettercase—understanding that successful font selection requires taste, training, and contextual nuance that AI needs to learn.
The Real Reason I Haven't Launched Yet
Honest answer: Survival mode is real. Freelancing's been tough. Finding stable work's sometimes challenging. It can be hard to focus on passion projects that don't immediately generate income.
Plus, let's be honest — ADHD is a double-edged sword with projects like this. Great for the initial obsession and prototype building, terrible for sustained focus. The cycle became: build something exciting, get pulled into client work, forget about it for months, then rediscover my notes and think 'oh right, this could actually help people.'
And, frankly, it's difficult to build an entire platform by yourself. Even when you're good at all the roles—product manager, designer, developer, sometimes marketer—it's still hard to do all of it yourself and get to a good enough finish line that it's ready to launch. There's a difficult balance between charging too little and people thinking little of your project, or charging too much and feeling like you need to build more to make it worth that price.
But here's what I've learned: I could've launched a mediocre font organizer years ago. Instead, I waited for technology to catch up to my vision.
And things are finally starting to come together — AI sophistication, modern development tools, cloud infrastructure, my skills and perspective and abilities.
What This Would Mean After 16 Years
Most people hear this idea and think "seems neat, I guess." When I pitch it to friends or family or other people who aren't designers, they don't quite get how it could be truly helpful to a designer's process.
I have a gut feeling deep down that it's one of those products that once it becomes part of your process, you won't be able to imagine living without it. Like how we used to manually organize photos before Google Photos started auto-tagging faces.
If I can build something that truly helps designers find the right fonts for their projects—something they use and care about — prove that this vision I've protected for 16 years was worth holding onto — that would be incredibly vindicating. And deeply exciting.
The Vision I'm Building Toward
Instead of showing you 10,000 fonts, show you the 3 fonts that actually make sense for your specific project.
I'm working toward a system that would include AI agents that understand project context ("tech startup, enterprise audience, trustworthy but approachable"), social discovery from designers whose taste you respect, a learning system that gets better at understanding your preferences, team collaboration for design system consistency, and a lot more.
The Lesson: Great Ideas Wait for Their Moment
Sixteen years taught me that the best ideas often arrive before the technology exists to execute them properly. I could've forced a decent font organizer to market in 2009, but instead I waited for convergence: AI sophistication, modern development tools, mature web font ecosystem, social platform patterns.
The pieces are finally aligned. Sometimes the most important product skill is patience—knowing when an idea is worth protecting while you wait for technology to catch up.
Lettercase represents 16 years of waiting for technology to catch up to a vision: AI agents that understand design context well enough to recommend the perfect typography for any project, not just fonts that look similar.