Team BitterSweet

Every major brand faces the same problem: they've built great products, but they don't know how to make them go viral. They're spending millions on TikTok marketing, yet most of that budget is burned on guesswork. The marketing teams we work with used to spend 8+ hours a day watching TikTok, scrolling through hundreds of videos, trying to reverse-engineer what made certain content take off. Then they'd brief creators, run campaigns, and hope for the best. Most of the time, the results were disappointing.

We believe the "intuition" behind virality can be learned by AI. Not by generating prettier videos — tools like Nano Banana Pro and VEO 3 already do that well. But by understanding the deeper patterns: what hooks work for which audience, what messaging drives action, what creative formats are trending right now. This is the missing layer.

So we're building a viral engine. It researches TikTok trends in real-time, tests combinations of audience, hook, and message with AI-generated content, stacks success and failure data, and trains models that get smarter with every campaign. The more brands use us, the better our AI becomes. Customer #20 gets better results than customer #1 — with less iteration.

Our early customers are seeing it work. Beauty brands running $10M+ monthly marketing budgets are using our system to find winning content formulas faster. One agency cut their trend research time by 90% and started publishing daily reports instead of monthly ones. A brand founder told us: "If you can test content ideas at $1 per variation before we spend $50K on production, that changes everything."

We think this is a $100B problem. The S&P 500 companies alone spend trillions trying to capture consumer attention. If we can solve even 1% of the "how do I make this go viral" problem, we're building something massive. TikTok is just the beginning — the same AI can work across Instagram, YouTube Shorts, and whatever platform comes next.

CEO (Junseong Kim) spent 7 years as an ML Researcher & Engineer at ScatterLab, Korea's #1 AI entertainment startup, building conversational AI before transformers existed. He also served as CEO Staff and Strategy Manager, helping scale the company from $0 to $20M ARR.

CTO (Sunghun Max Kim) built the LLM data flywheel pipeline at ScatterLab and the core payment infrastructure at Korea's largest travel company. He's obsessed with data systems that compound over time.

Together, we've won Europe and Asia's biggest hackathons (Junction, Junction ASIA) twice in a row. We know how to build fast and ship under pressure.