AI & Machine Learning
The shared problem is trust: when a model scores a sales call, reads a biosignal, or calls a deal at risk, people need to understand what it saw, why it surfaced the signal, and when to overrule it. These projects all center on that problem: a CRM-native intelligence product that shipped, a research platform run by AI agents, and a concept that put task-scoped agents on screen two years before the industry settled on the term.
Intelligence products I created end-to-end and owned through launch, where model outputs had to be legible enough for someone to act on them.
I led design strategy for conversation intelligence inside Sales Cloud: recorded calls and meetings turned into comparable coaching signals (talk-to-listen ratio, competitor mentions, objection types), anchored to the CRM record.
View case study
A research-grade biosensor ring, a participant app run by AI agents, and an analyst platform that turns raw biosignal into moment-by-moment insight into unconscious reactions to media.
View case studyExploratory work that drew a system of task-scoped agents two years before the industry settled on the term.
With the Slack acquisition announced and every product leader asking what it meant for them, I led a four-person team to answer the question in pictures: a full deal-save narrative putting Slack at the center of the seller’s day.
View case study