
When Google released Gemini 3.0 in November 2025, most companies took weeks to figure out how to incorporate it into their operations. Sekai, an emerging social platform built for the AI era, took a different approach: the founding team pulled two all-nighters and shipped an AI coding agent 48 hours later.
That sprint set off a chain reaction, and within one month, the no-code AI platform scaled from zero to 50,000 daily app creations. Now, by early 2026, Sekai is on track to surpass two million mini-apps within three months of launch, matching the total number of apps Apple’s App Store has accumulated over more than 17 years of operation.
The company calls itself “TikTok for software,” and the comparison is more than marketing. Sekai’s founder Lucky, a former Google engineer who previously built and sold platforms to Apple and ByteDance, spent a decade studying how content formats evolve from professional tools into social entertainment.
His thesis is straightforward: as creation costs approach zero, the future of virality will shift from videos to shareable mini-apps, and Sekai aims to lead that transition.
Sekai’s Goal: Turning Ideas Into Apps
Traditional app development demands thorough technical knowledge, not to mention months of work and access to significant capital, yet millions of people carry creative ideas they can’t execute. Sekai appeals directly to that gap. “We’re targeting the 99% of people in society who wouldn’t call themselves builders, who wouldn’t even think developing apps is a thing for them,” founder Lucky said.
The platform works by converting a prompt into a functional mini-app. A user types a description, and Sekai’s AI coding agent generates the application within seconds. Someone using the app could create, for example, a Christmas card that only opens after midnight on December 25, or a meme that roasts a specific friend by name.
Sekai now processes 50,000 daily app creations, reaching 50% of the volume produced by Lovable, the current market leader that recently announced a six billion dollar valuation. “We’re still growing very quickly,” Lucky said, and he points to the app’s accessibility as a core selling point. “A lot of people may not be very tech-knowledgeable, but they do spend a lot of time going through social media and become naturally inspired. This app allows them to bring that idea to life.”
Sekai’s thesis points to a historical pattern where content formats change once they become more accessible to the public. As Lucky puts it, SMS texts evolved from email, and video is currently shifting from movies as its mainstream medium to TikTok. Software, Lucky argues, is next in this progression. “Software will go through the same thing,” he said. “For the past 40 years, software has been designed to serve utility, meaning it has to solve a problem. But once everyone can create it, you’ll have billions of software.”
The apps emerging on Sekai reflect this transition playing out. Users create disposable, shareable applications for self-expression rather than problem-solving. The creations themselves serve no lasting utility beyond entertainment, and that’s precisely the point.
Following The TikTok Playbook
One core aspect of Sekai’s creation and development was the founder’s interest in how TikTok became the dominant social media platform it is today. When TikTok first emerged, it spread not through mass marketing but through social clusters: one student in a classroom would start creating lip sync videos, and the app would quickly spread throughout the rest of the class.
Sekai’s AI-native social platform is following the same pattern. “One person starts playing with it,” Lucky said, “and quickly the whole group starts playing with it.”
Lucky and his team point to ways in which they can accelerate this growth. For example, the platform focuses on mini-app creation and not full websites or SaaS tools. This limitation mirrors TikTok’s early discipline of restricting videos to exactly 15 seconds, a constraint that forced creators to have their ideas fit the format rather than simply porting content from other platforms.
Additionally, Sekai’s recommendation engine, built by a former TikTok tech lead with nine years at ByteDance, applies algorithmic curation to the platform’s mini-apps. The system surfaces creations that work well, filters out broken or complex outputs, and identifies which formats are most shareable and remixable. Plus, users can modify existing mini-apps like DJs remix tracks, with customized controls for each content type.
Speed as Strategy
Sekai had been operating for over a year before the Gemini 3.0 launch, building interactive text-based experiences for anime fans and gamers who were drawn to immersive, fictional worlds. The platform had accumulated a decent amount of users, but the content remained niche and difficult to share. When Google released Gemini 3.0 in November 2025, Lucky recognized an inflection point.
“On the day we saw it, we tried it, and we realized that this could increase our app’s capabilities beyond what we had previously imagined,” he said.
Lucky and his team pulled two consecutive all-nighters to ship a coding agent within 48 hours of the Gemini release. By using this generative AI tool, the sprint allowed Sekai to begin processing user-generated mini-apps before competitors could react.
That urgency paid off. Less than two months after launching the AI mini-app creation feature, Sekai hosts nearly one million apps, a milestone that took Apple’s App Store 17 years to reach, and Lucky only expects that number to continue growing. “Given our speed, we can reproduce the whole App Store’s apps within three months,” Lucky says. “That’s how fast the world is changing.”
A New Way To Build
By giving non-technical users access to generative AI tools that create functional mini-apps in seconds, Sekai is transforming software from a professional tool into a way for regular people to express themselves. The platform’s trajectory seeks to position the next wave of viral content to shift away from videos and into playable, interactive applications built by people who have never written a line of code.
Disclaimer: GeekWire newsroom and editorial staff were not involved in the creation of this content..