In this raw and personal episode of Coffee with a Founder, Feifan Wang sits down with host Beck Bamberger to discuss the human side of building SourceMedium. While Feifan is known as a data expert, this conversation peels back the layers to reveal his journey as a non-technical founder who immigrated to the U.S. from China at 14 without speaking English—an experience that forged the resilience needed to weather the startup storms.
Feifan opens up about the “second greatest tech crash of all time” in 2022, the pitfalls of scaling a team too fast after raising seed money, and his radical pivot to a “lean and happy” philosophy. He explains how he restructured SourceMedium to prioritize automation and AI leverage over headcount, allowing a small, elite team to outpace competitors 5x their size while maintaining personal happiness and peace as primary KPIs.
Key Takeaways
- The Immigrant Edge: How being “air-dropped” into 7th grade in Maryland with zero English skills taught Feifan the ultimate entrepreneurial skill: adapting to total uncertainty and surviving.
- The “Resident” Origin Story: Feifan’s path from business major to VP of Product at a $1B exit (Resident Home), where he learned that managing data is a universal pain point, whether you’re selling mattresses or SaaS.
- Lean Team, Huge Output: The decision to shrink from 15+ employees back to ~5, using AI and automation to “100x productivity” rather than solving problems by throwing bodies at them.
- AI as a Force Multiplier: How Feifan uses AI personally to code faster and build products that would previously require a massive engineering team, turning “team size” into a creative constraint.
- Happiness & Peace as KPIs: A refreshing take on founder mental health—rejecting the “grind until you die” mentality in favor of building a sustainable, fulfilling business that serves the founder’s life, not the other way around.
- The “Slack Analyst” Vision: Fulfilling a 5-year-old vision of an AI agent that lives in Slack and answers data questions 24/7—something that was impossible in 2020 but is now a reality with LLMs.
