Project Type:
UX Research & Design.
Year:
July, 2025
KOSI.AI
about.
Kosi is a unified product insight hub designed to help teams make smarter product decisions by bringing all their research together in one place. Its AI-powered UX research assistant focuses exclusively on qualitative data interviews, notes, usability tests, and open-ended feedback synthesizing them into clear, structured insights without touching quantitative analytics. Instead of automated behavioral analysis, Kosi empowers researchers with full control, transforming raw qualitative inputs into personas, journey maps, themes, and narratives that preserve the human lens of UX research.
challenge.
Product teams struggle to turn scattered qualitative research into clear, actionable insights. Existing tools either overwhelm researchers with complexity or over-automate the process, reducing control and compromising the human context behind user behavior. Teams need a way to synthesize interviews, notes, and feedback with clarity and consistency while preserving the researcher’s judgment and expertise.
design.
The design process began with conducting user interviews to understand motivations, frustrations, and expectations around ordering food through an app. These insights helped uncover real pain points and shaped the direction of the solution. Based on what I learned, I moved into the ideation phase, creating rough sketches and low-fidelity wireframes to explore different layouts and interaction patterns. This stage was focused on rapid iteration and validating the overall structure before committing to visuals. Once the flow and hierarchy felt solid, I transitioned into high-fidelity designs, refining UI elements, visual styles, and overall usability to deliver a polished and intuitive final experience.













