I find the real problem behind messy work, then build the system that solves it.
I use research to understand people quickly, product thinking to find the right problem, and AI to make repeated work faster when the workflow around it is designed well.
Start with the problem. Use AI only when it helps.
Human struggle comes first
I start by understanding what people are trying to do, where they get stuck, and what would make their work easier.
AI is a tool, not magic
I do not start by asking where to add AI. I start with the problem, then decide whether AI is the right tool.
Good AI systems design around limits
AI has strengths and weaknesses. A good system uses the strengths and adds structure, checks, and human judgment where the model is weak.
Start small, learn fast
I prefer shipping a small useful version, watching how people use it, and improving quickly instead of designing a huge AI product upfront.
UI Prototyping Tool
I designed and tested a Miro workshop tool so Salesforce teams could build Service Cloud screens with customers in live sessions.
Read case study →Cultural Awareness Toolkit
I researched cross-cultural teamwork and designed a card game workshop that helps teams talk about working styles early.
Read case study →Have an unclear customer problem or manual workflow that could work better?
lucyfengcl@gmail.com →