Master's Research - Design Tacit Knowledge Capture System

Capturing the thinking behind the interface.

Design decisions often happen without being written down. My master’s research explored whether Figma interaction data could help document this implicit knowledge without repeatedly interrupting a designer’s work.

I developed and evaluated a prototype that turns interaction traces and task context into candidate knowledge for designers to review.

AI proposes. The designer decides.

The workflow groups Figma actions into design episodes, then uses an LLM to propose short knowledge cards about possible intent, strategy, and reasoning. These are interpretations to question—not definitive accounts of a designer’s thinking.

A workflow built around expert review.

I combined interaction logs, task context, episode-based processing, and LLM-generated cards in a prototype workflow. Professional designers reviewed the cards through keep, edit, or discard decisions.

The study compared this material with retrospective think-aloud reflection to examine the usefulness and accuracy of the generated knowledge.