ProductComing soon
How Multi-Agent Systems Design a Day: Orchestration Notes
The difference between a plain LLM call and multi-Agent orchestration. Sijo, folk song, pansori: the design decisions behind an Agent family composing one person's day.
Writing
We write as we build.
We publish what we learn while building DailyFit. How we run a company as a team of AI Agents, our hypothesis about the 55+ market, and even the things that did not work out.
Rather than finished conclusions, we try to leave behind the thinking that led to them. We believe making public what we believe, and where we were wrong, builds a better company. We call this Radically Transparent.
Youngwoo Michael Suh, Founder · DailyFit
Essays1 published
OperatingComing soon
Running a Company with a Team of AI Agents
AI Agents split strategy, research, product, and design; people make the judgment calls. The real operating structure that lets a small team move with the speed and rigor of a large one.
ThesisComing soon
The Korean 55+ Market Thesis
15 million people, the generation moving into digital the fastest. Why this market is the opportunity of the next decade: AI is the means, the 55+ generation is the identity.
BuildingComing soon
To Beta: What a Small Team Cut to Ship Every Week
The decisions we deliberately chose not to pursue for the sake of speed. Why narrowing priorities turned out to be harder than adding features.
DataComing soon
What English LLMs Miss: Collecting Korean 55+ Language Data
Why we collect and refine the data ourselves. How tone, phrasing, and context become a real moat for the model, and the practical difficulty of the work.