Home

AI PRODUCT DESIGN

I design AI products that people actually use

Most AI features get ignored. I've spent years figuring out why, talking to hundreds of users, and shipping products that break that pattern. Currently Senior Manager of UX Design at Counterpart Health.

 
 

Focus

LLM-powered products

Previously

Filevine, JPMC, Nationwide

Role

Design leadership + IC

 

POINT OF VIEW

What I've learned designing AI products

Users don't care about AI

They care about getting work done. The best AI products disappear into existing workflows. I've watched "revolutionary" features get ignored because they asked users to change how they work. The wins come from embedding intelligence where people already are.

Trust is designed, not assumed

When an AI gives a wrong answer, users don't blame the model. They blame the product. Source attribution, confidence signals, inline feedback loops, and graceful fallbacks aren't features. They're table stakes for AI that people rely on.

Ship the boring version first

AI Fields that auto-analyze uploaded documents. An existing chat app upgraded with an LLM. A doc review action embedded in a context menu. None of these are flashy. All of them shipped and stuck because they solved a real problem in a familiar shape.

 

Selected work

AI and product design at scale

AI · LLM · EMBEDDINGS · LEGAL TECH

Legal AI Chat

65%+

reduction in document review time

Redesigned a peer-to-peer chat app into an LLM-powered legal assistant. Users query case data and documents in natural language, with proprietary embeddings driving accurate, cited responses.

Read the case study →

 

PAYMENTS · 0→1 · STRIPE · B2B

Enterprise Payments Platform

Designed the full payment lifecycle from scratch, integrated across an app ecosystem. Built for accountants and managing partners at law firms who needed to send invoices and collect payment without leaving their tools.

75%

faster time to payment

 

SYSTEMS · ACCESSIBILITY · SCALE

Cross-Product Design System

Built the design system backbone for an ecosystem of legal tech products. WCAG-integrated component library that enabled consistent, accessible experiences across multiple applications and development teams.

Read the case study →

 

BACKGROUND

12+ years across enterprise and startups

Counterpart Health
Sr. Manager, UX Design
2025 – Present

Filevine
Director of Product Design
2021 – 2025

J.P. Morgan Chase
Manager, UX Design
2019 – 2021

Nationwide
UX Engineer
2013 – 2019

Abercrombie & Fitch
Senior UX Designer
2017 – 2018

 

HOW I WORK

Approach

Start with the workflow, not the technology

I map how people actually work before proposing where AI fits. This means sitting with users, watching them skip features, and finding the moments where intelligence could reduce friction without adding steps.

Build trust through transparency

Source citations, inline feedback, confidence indicators, and clear fallback paths. Every AI interaction I design gives users a reason to believe the output and a way to correct it when it's wrong.

Iterate with data, not assumptions

I validate designs through user testing and track adoption metrics post-launch. If something isn't getting used, that's signal. I've killed features I designed because the data said they weren't working.

Build teams that ship fast

High iteration velocity matters more than perfect process. I build design teams that can work through ambiguity, collaborate directly with engineers and data scientists, and ship incremental improvements rather than waiting for big reveals.

 

Let's build something
worth using

I'm open to conversations about AI product design roles, consulting, and advisory work.

Send me an email or connect on LinkedIn.