Understanding user intent
Investment search is rarely precise. Users express goals, not taxonomy: “low risk,” “green,” “dividend income.”
The first step was identifying how natural language could be translated into structured financial filters without overwhelming the user with complexity.