top of page

SCATTER LAB: Can AI Be Personal, Emotional—and Profitable?

For over a decade, ScatterLab has pursued a simple but radical question: Can artificial intelligence feel more human—not just functional?

Long before large language models became mainstream, Kim Jong-yoon, a business and sociology student at Yonsei University, was already exploring how language reveals emotion. His university research project—analyzing the correlation between text messages and romantic interest—won a government startup grant in 2011. That project became the foundation for ScatterLab: a company focused not just on AI, but on emotional AI.

 

ScatterLab’s early products, like Text At and Science of Love, were playful tools that helped users interpret conversations and understand romantic signals. While light in tone, these apps quietly built something serious: a large, high-quality dataset of natural Korean conversations, and a philosophy of AI that treats human nuance as a core design challenge, not an edge case.

The company’s journey hasn’t been without setbacks. In 2020, it launched Lee Luda, a conversational chatbot modeled after a 20-something friend. The service went viral—but was soon suspended due to ethical issues around language moderation and data handling. For many companies, this could have been the end. But for ScatterLab, it became a turning point. The team took a hard pause, rebuilt its data pipelines, added multi-layered safety reviews, and recommitted to responsible deployment.

That learning process culminated in Zeta, launched in 2024. Unlike most AI chatbots, Zeta doesn’t aim to answer questions or replace tasks. Instead, it’s a storytelling platform—where users co-create characters, build relationships, and immerse themselves in emotional narratives. In Korea, Zeta reached 2 million users within its first year, with exceptionally high engagement: over 2.5 hours of average daily usage. A Japan launch followed quickly, and for the first time in its history, ScatterLab reached break-even.

At Saehan Ventures, we saw in ScatterLab not just a technical AI company, but a rare product intuition: the ability to turn raw technology into cultural experience. We invested at an inflection point—after Zeta’s initial breakout, but before its global expansion. We believed the team had something few in the AI space possess: a deep understanding of human behavior, and a willingness to learn through failure.

AI is entering every industry. But we think some of the most enduring companies will come from those who treat it not just as computation—but as conversation.

bottom of page