01 The strangest lead source we ever had
A one-day app, and the CEO who saw it
In 2022, before AI made speed fashionable, we ran a public challenge on Twitter: could we ship a working mobile app, admin panel included, in a single day? We live-threaded the build through the day, design to deploy, and by the evening the app existed and worked. It was a stunt, but an honest one, and it traveled further than we expected.
Alican Karaman, CEO of Mixer Games, saw that thread at exactly the right moment in his company's life. He wrote to us, introduced himself, and laid out the situation with unusual honesty for a first message: the state of their funding, the state of their runway, and a project called Fundle that had to exist very soon. The company had built games people enjoyed. What it needed now was the app to carry them all, and time was the one resource it did not have.
02 A cornered CEO trying to pull a rabbit out of a hat
Months of runway, no app
The situation was blunt. Mixer Games had perhaps five or six months of runway left, maybe less. The company's next chapter depended on Fundle, a mobile app bundling dozens of their games into a single library, and it did not exist yet.
Mixer had a real in-house team, but Alican refused to split its focus away from the work already in motion. His reasoning was disciplined: pulling the team onto Fundle meant losing momentum everywhere at once, and the math said an internal build would not land inside the runway anyway. His own words later became the summary: doing it internally would have meant hiring at least six new people and having them adapt instantly, which is not how hiring works.
In the first meeting we met a CEO who was cornered and knew it, calm on the surface, looking for someone to pull the rabbit out of the hat. We said yes, and the clock started on August 19, 2022.
03 Flutter would only carry one Unity project. The app needed dozens.
The engine problem
The hardest problem surfaced mid-build, the kind you only discover by building. Fundle was a Flutter app hosting games made in Unity, and Flutter's Unity integration allows exactly one embedded project. An app whose entire premise was many games living in one place was blocked by its own toolchain, and no amount of configuration lifted the constraint.
So we went around it. Instead of embedding Unity projects, we exported the games as WebGL builds and rebuilt the app as a fast, seamless host for as many of them as the catalog needed. Fundle's team optimized on the game side, we optimized the server and app layers, and the result ran smoothly enough that players never sensed the workaround underneath. A constraint that could have sunk the timeline became an architecture decision that made the catalog effectively unlimited.
- Flutter app with a WebGL-based multi-game architecture
- Server and app-side performance optimization for smooth play
- Design built jointly with the Mixer Games team
- Admin and content tooling to manage the game catalog
04 Dozens of games behind a single install
One app, a whole arcade
For players, Fundle is deliberately simple: one install, a whole library. Open the app, browse the catalog, tap a game, and play, with no further downloads, no separate installs, and no friction between one game and the next. For a studio like Mixer, that simplicity is the entire business case. Instead of fighting for attention with a dozen separate store listings, every game they make lands directly in front of an audience that already opened the door.
Behind that simplicity sits the WebGL architecture doing quiet work: games load fast, run smoothly, and new titles can be added to the catalog through the admin tooling without touching the app itself. The catalog can grow as fast as the studio can make games, which is exactly the property a game library company needs from its platform.
05 Speed is subtraction, not heroics
How six and a half weeks actually works
Speed like this is not overtime and luck. It is subtraction. There was no discovery theater and no workshop calendar. Scope was settled in short working sessions we drove with concrete proposals: the client said what outcome he needed, we brought the product decisions, and the meeting count collapsed. Days that would normally disappear into alignment went into the build instead.
The split of labor stayed clean throughout. The games were Mixer's, the design was a genuine collaboration with their team, and everything else, from architecture to the admin panel to release, was ours. Testing ran continuously alongside the build rather than as a phase bolted on at the end, so the final week was a finish line, not a firefight. Six and a half weeks sounds like a sprint. From the inside it felt like a straight line.
06 Handover in six and a half weeks, a raise a year later
The outcome
We handed Fundle over on October 3, 2022, six and a half weeks after kickoff, with 10 to 12 games live in the app. For a while afterward we provided maintenance and additional development, and then passed the product cleanly to Mixer's own team, which is exactly how a rescue engagement should end: the client keeps the product, not the dependency.
About a year later, the news came through. Mixer Games closed a new investment round at a $5M valuation, raised on the strength of the company that Fundle kept alive. Alican's words are the ones we keep: without you we could not have done this, and we would have gone under.
Startups live and die by timing. This one lived long enough to fight the next round, and that was the entire assignment.
And the origin still makes us smile. A one-day app challenge on Twitter, posted for the fun of it, turned into the project that kept a funded startup alive. Proof of speed travels further than any pitch deck we could have written.
