my first real project. built a roblox game that somehow blew up, even though i couldn't spell "chaotic" correctly. it peaked at 20k players and 223k combined views. not bad for a 9 year old
the spark
i was obsessed with Jailbreak. at its peak it hit 530k players online at the same time. but for me it got repetitive fast, you robbed the same spots and did the same things over and over. i kept thinking: what if i built my own version, but better?
step 1: building the team
i found a few people on a discord server who liked the idea, pitched them, and put together a small team: coders, a designer, a modeler. i was 9, paying each of them in roblox currency, a tiny entrepreneurship project before i even knew the word.
step 2: a beta in silence
we built quietly for a few months. no announcements, no hype, just a beta we kept polishing until the game actually felt fun. i added way more robberies than jailbreak had, and a brand-new role: the citizen. instead of only cops and robbers, you could roam the city, feed criminals clues, or team up with the police. it made the world feel alive instead of a loop.
step 3: getting the youtubers
then i started reaching out to jailbreak content creators to show off the game. it worked, across all their videos, caotic robbers pulled in 223k combined views, and our own trailer hit 11k views. for a game made by a 10-year-old and his discord friends, that felt unreal.
step 4: someone deleted everything
one year of work. one month to rebuild it.
a month before release, one of the developers decided he wanted to be the leader. i said no. he wouldn't accept it, and because everyone had full access in roblox studio, he deleted the entire map. a year of work gone in seconds, and we had no backups.
so we rebuilt the whole thing in one month, scrambling, cutting corners, prioritizing what mattered and skipping the rest.
step 5: the launch (a disaster)
here's the cruel part: the youtubers had filmed the old game, the polished version from before the deletion. so players showed up expecting that, and got our rushed one-month rebuild instead. it didn't match, and it showed.
we still hit a peak of 20k players. but the game wasn't ready, and a few days after launch we had to take it down.
what it taught me
two technical lessons i never forgot: give people access to only what they actually need, and always keep backups. but the bigger one was about ownership, i told the team “this was my mistake,” because it was my fault for giving everyone permissions. owning that, instead of blaming him, was the most valuable thing i took from the whole project.
QuiverAI, building AI tools for vector graphics. I'm working my ass off to build the next-generation platform for designers.
HackSpain
