I build AI systems
that solve real problems.

Most people use AI as a tool. I design AI-native workflows — systems where multiple AI agents handle entire production chains end-to-end, from decision-making to execution to self-review. My content pipeline runs six specialised agents that took a debut Douyin video to 76,000+ views and consistently hit a 14.3% save rate (vs. 3% industry average). That's not prompting — that's systems design.

I'm an M2 economist at Panthéon-Sorbonne with a dual degree in Finance from UIBE Beijing. Economics gave me an empirical mindset — and right now, getting the most out of AI is itself an empirical science. You try things, measure what works, push boundaries, analyse the data, and iterate. That's how I approach everything: dispatch optimisation in Python, content hook experiments, or designing agent chains.

I work across four languages and three continents. I've coordinated 100+ volunteer programmes (including the 2022 Beijing Winter Olympics) and built a digital platform to 2,000+ subscribers. But what I'm really doing is building toward a future where one person with the right AI systems can outproduce an entire team — and I'm proving it with my own work first.