Agentforce for Setup
Designed the agentic layer that lets admins configure Salesforce through natural dialogue instead of nested menus: an AI agent acting on production systems, built around trust, control, and intent.
I design and build products that keep human judgment at the center of AI, helping people understand what's happening, stay in control, and trust the systems they rely on.
I'm a product designer with a background in psychology and human-computer interaction design who designs intelligent systems. Over the past decade, my work has evolved from identity and trust experiences to enterprise AI products, where I help people maintain agency, build confidence, and work effectively alongside AI as it takes on more.
Today, I'm a Lead Product Designer at Salesforce, designing AI-powered experiences for the people who configure, automate, and maintain the systems businesses run on. I also build with AI directly, taking ideas from rough concept to working prototype.
Outside of work, I'm a mom of two young kids. Parenthood has reinforced many of the lessons that guide my work: people are unpredictable, trust is earned, and the best systems create space for others to grow.
Every chapter here is a version of one problem: as AI takes on more of the work, how do people stay oriented, in control, and able to trust what a system is doing on their behalf? Agentforce Coworker was the turning point, the project where answering that became the work I most want to do.
Four chapters, one throughline: turning powerful, complex systems into experiences people can trust.
Designed the agentic layer that lets admins configure Salesforce through natural dialogue instead of nested menus: an AI agent acting on production systems, built around trust, control, and intent.
Built a knowledge agent for Salesforce’s whole builder ecosystem, end to end, as a designer. Its “Ask the Hub” agent routes you to the right tool, synthesizes cross-team research into one view, and audits a PRD against our trust and autonomy principles. I cleared the full enterprise AI governance path myself. It’s now under exec review for adoption across our UX org.
Led the setup & configuration experience for a new agentic search product that unifies data across Salesforce CRM, Slack, Data Cloud, and third-party sources, turning a fragmented ecosystem into one searchable brain.
Unified twelve fragmented product logins into a single, accessible email-based sign-in and environment switcher: the front door every Salesforce user passes through.
Redesigned the framework developers use to connect external apps to Salesforce via OAuth, SAML, and OpenID Connect: V2 of Connected Apps, unifying a scattered setup into one connected, local-by-default experience.
Redesigned how the Salesforce Authenticator app explains its automated logins, making the language and signals behind why an automation succeeded or failed clear to millions of mobile users.
When I joined Salesforce's Builder team, I used AI to ramp up quickly on a complex product spanning no-code, low-code, and pro-code experiences. Along the way, I built AI-powered tools for the team, created a working editor prototype that informed engineering's implementation, and demonstrated how designers can move beyond prompting to building.
Each team site updates itself with Claude Code and Slack MCP scripts; the hub pulls from all of them, refreshing on every run.
A knowledge agent for the whole builder ecosystem. Its "Ask the Hub" agent routes you to the right tool, synthesizes cross-team research into one view, and audits a PRD against our trust and autonomy principles. I built it end to end and cleared the full enterprise AI governance path myself, as a designer.
Read the case studyA hub for my new team, plus a separate site and repo for every team at our builder summit, each handed off individually. Scheduled Slackbot skills keep each site's resources, design files, and research current on their own, with a live Slack feed that refreshes on every run, and the Builder Hub pulls from all of them.
My /hybrid-editor-prototype skill, scaffolding a complete working prototype from one command.
A Claude skill I built (with a starter site) for prototyping the new hybrid editor. One command scaffolds a complete working prototype to build from instead of a blank file, plus copyable components and curated resources, so anyone goes from zero to running in minutes.
An agentic dashboard for tracking AI builder tools across the industry: compare them side by side, audit new ones worth watching, and generate reports for different audiences. I laid out the UI in claude.ai/design, then handed that file to Claude Code to build it out. A scheduled Slackbot skill keeps it current, scanning our UX and builder channels each week and sending the team a Monday digest.
A suite of ten Claude skills that runs my design practice as a system: capturing a build as it happens, drafting the case study in my voice, trimming and critiquing the work, and pre-flighting the site before I share it. A retro skill keeps sharpening how I prompt. I built them while porting this site over, and handed them one case study to draft end to end. They're open on GitHub.
Built by a skill in the suite, rendered in my own design tokens. Flip the Classic / Expressive / Editorial toggle (bottom-right) and the diagram re-themes itself live.
Most of these are internal tools I built to onboard myself and support my team, so the details stay light here. The Claude skill suite is public and on GitHub. Full narrated walkthroughs available on request.
Intelligent systems earn trust through clarity, transparency, and control. I design experiences that make AI behavior understandable, predictable, and reversible before users commit to action.
Technology should make people more capable, not less. I design systems that keep people informed and in control, helping them accomplish more than they could alone.
My work often lives in complex systems, from identity to enterprise AI. I transform ambiguity into intuitive experiences that help people navigate complexity with confidence.
The best solutions emerge through curiosity. I seek to understand people, systems, and emerging technologies deeply before shaping what comes next.
These principles have guided my work from identity and enterprise search to agentic workflows and AI-powered creation tools.
I'd love to hear from you.