Salesforce Enterprise AI Interaction Patterns
A UX Research Project with Salesforce and UC Berkeley
Problem
Enterprise users often struggle to understand, trust, and effectively use AI-powered tools due to poor onboarding, unclear workflows, and a lack of control.
Approach
We conducted a mixed-methods study including surveys, user interviews, competitive analysis, and usability tests to explore how AI features affect learnability and user control across roles.
Solution
We identified key friction points and recommended UX strategies such as layered onboarding, contextual feedback, and override options to help AI tools fit more naturally into real-world workflows. The results were published in a comprehensive slide deck and presented to the Salesforce team.
This project was part of a semester-long graduate course at UC Berkeley’s School of Information, conducted in collaboration with Salesforce as our industry client. The research was supervised by Steve Fadden, Google Cloud’s UX Research Lead and lecturer at UC Berkeley.
Salesforce Enterprise AI Interaction Patterns
A UX Research Project with Salesforce and UC Berkeley
Problem
Enterprise users often struggle to understand, trust, and effectively use AI-powered tools due to poor onboarding, unclear workflows, and a lack of control.
Approach
We conducted a mixed-methods study including surveys, user interviews, competitive analysis, and usability tests to explore how AI features affect learnability and user control across roles.
Solution
We identified key friction points and recommended UX strategies such as layered onboarding, contextual feedback, and override options to help AI tools fit more naturally into real-world workflows. The results were published in a comprehensive slide deck and presented to the Salesforce team.
This project was part of a semester-long graduate course at UC Berkeley’s School of Information, conducted in collaboration with Salesforce as our industry client. The research was supervised by Steve Fadden, Google Cloud’s UX Research Lead and lecturer at UC Berkeley.
Salesforce Enterprise AI Interaction Patterns
A UX Research Project with Salesforce and UC Berkeley
Problem
Enterprise users often struggle to understand, trust, and effectively use AI-powered tools due to poor onboarding, unclear workflows, and a lack of control.
Approach
We conducted a mixed-methods study including surveys, user interviews, competitive analysis, and usability tests to explore how AI features affect learnability and user control across roles.
Solution
We identified key friction points and recommended UX strategies such as layered onboarding, contextual feedback, and override options to help AI tools fit more naturally into real-world workflows. The results were published in a comprehensive slide deck and presented to the Salesforce team.
This project was part of a semester-long graduate course at UC Berkeley’s School of Information, conducted in collaboration with Salesforce as our industry client. The research was supervised by Steve Fadden, Google Cloud’s UX Research Lead and lecturer at UC Berkeley.