Evaluating the voice AI experience of LG ThinQ On

I worked with LG Electronics to understand how people use voice commands at home and develop a way to evaluate the LG ThinQ On’s responses. Since the speaker was still in development, we asked participants to record everyday requests and used a GPT-based voice prototype to explore what they expected in return.

I helped plan the research, review related studies, and group findings into common themes. We used what we learned to create evaluation criteria and review conversations from lab testing, looking at whether the speaker understood the request, responded appropriately, and gave clear answers.