Planning the Path to a Better AI Assistant

When a global company wanted to improve the frequency of use and quality of interactions with its AI assistant and develop new uses when given the ability to “see” and “hear” what a user was seeing/hearing, I developed an online diary study followed by observed product use with a smaller subset and then 1:1 interviews to uncover current motivations, triggers, abilities of users and their inflection points, and discover future uses.

Then, I designed a path forward for the company in 4 tiered stages with metrics for success at each stage–including novel ways to teach both the user and the AI how to interact with each other and improve the trust of the AI’s quality of response at each stage. The company was delighted and is currently implementing that path.