I design both sides: what people see, and what the people doing the work use.

Today I lead product design for Instagram Shopping as an IC. Before that I was the first designer at Nuro, and the second at Instacart. I’ve managed teams of up to seven designers, and studied engineering and HCI at Stanford.

Instagram Shopping · 01

Design lead · 2020–2023, 2025–now

Making shopping work inside an app people don’t open to shop.

We tried building a shopping destination. It didn’t work. People open Instagram to be entertained, then shop when content inspires them. The job is to make that moment easy to act on.

Shoppable formats

We built shopping into the formats instead of a place to go.

Feed, Stories, Reels, and Live each carry products differently. I managed the team across all four.

I’ve worked on Shopping twice. First on seller supply, as an IC and then a manager. Now I’m focused on making it easy to shop anything you see.

Current scopeIC, setting direction across several teams

Suggested shopping search

We suggest the search instead of waiting to be asked.

A tap opens products and related media. A follow-up question narrows the results. The hard part was choosing how broad the search should be, and whether to suggest one at all.

Business AI

We let AI take the first business message.

Instagram’s first business AI. A prompt on a business’s story opens a private chat that clearly says it’s AI, so someone gets an answer before buying instead of waiting days for a reply.

Creator monetization

We opened tagging to any link, not just our catalog.

Creators can earn on products Instagram doesn’t carry, and place each card where it belongs on the timeline.

Seller onboarding

People could check out directly on Instagram.

Setup worked on a phone for the first time, on web, or through Shopify. Sellers only completed what they needed. In-app checkout went deeper into their payment systems so someone could buy in the moment.

Subscriptions · 02

Design lead · 2023–2025

Building a subscription worth paying for.

Meta Verified has one name but is really five products: a creator subscription and four business tiers, from $15 a month to about $500. Creators and businesses want different things, and every tier has to earn its price.

Why people paid

The benefits had to be things professionals couldn’t get anywhere else.

We focused on what professionals already paid other services to do: grow an audience, prove they were legit, run an account with a team, and protect their work.

It was Meta’s first subscription, so we built the basics from scratch. I led Instagram design and owned the benefits we launched. Facebook adopted them too. As the subscription grew, I began managing another designer while remaining primarily focused on IC work.

What happened nextIt worked well enough that Meta funded new subscription products off the back of it.

Live nowMeta Verified ↗

Links in content

Verification is what finally made links possible.

Instagram kept links out for years, partly because of spam and mostly to keep the app from turning into Facebook. Verified subscribers could finally add them, with limits on how often and how large they appeared.

Notability

The badge stopped telling you who was notable.

Once anyone could get verified, people checked follower counts to tell accounts apart, so we put the count next to the name. Creators could also filter comments and messages to other verified accounts.

Shared accounts

Messages and stories came from the real person.

Big accounts had been sharing passwords. Now owners could add people, see what each person did, and keep messages and stories private unless they chose otherwise. I set direction rather than doing the day-to-day design.

Links on profile

I designed Instagram’s answer to link-in-bio services.

People already paid outside services for this because a profile is where you decide if someone is legit. It became one of the two biggest reasons to subscribe.

Original credit

The subscription paid for the part that needed a human.

Big accounts copy smaller creators and keep the engagement. Detection ran for everyone, but subscribers could flag a post. If they were right, the copy was labeled with their name.

Nuro · 03

First product designer · 2018–2019

The vehicle was built, but the product wasn’t.

The team had a working vehicle and a strong self-driving stack. No one knew whether the product was an ice cream truck you walk up to, something you hail, or a fleet behind someone else’s app.

Customer handoff

With no driver, the vehicle had to explain itself.

Finding the right vehicle, opening the right compartment, and knowing when the delivery was done all had to be easy to understand on your own.

As Nuro’s first product designer, I worked across the customer app, vehicle interface, remote operations, and fleet tools. I also ran the field research behind the hardware experience.

PatentNamed inventor · US 11,623,657 ↗

Remote vehicle startup

One person could start the whole fleet.

Starting a vehicle used to require an operator standing beside it with a laptop. Now one person starts them all and sees exactly what blocks any car that fails.

Grocery service pilot

We ran the service before the vehicle could.

We hired drivers and shoppers and ran real orders. Sometimes a Nuro vehicle showed up instead of a car, so we could watch how people handled it long before the vehicle could do the whole job.

Fleet scheduling

We moved the fleet off Google Sheets and onto a real schedule.

We put every car, operator, and task in one calendar, from mapping runs to build tests. Operators got their day sent to them, and later the system began proposing the schedule.

Field research

How the vehicle greeted you mattered as much as any screen.

We had no researcher, so I ran sessions in daylight, at night, and in the rain. People were nervous approaching the vehicle, especially the doors, so sound, voice, and the screen had to explain what was happening.

Instacart · 04

Second product designer · 2015–2018

No one was designing the work behind the groceries.

When I joined, the customer side had a designer. The people doing the shopping had none, and in-store shoppers and drivers ran the whole operation on mobile web.

Fieldwork

I took shopper shifts myself.

That’s me. The product made more sense once I was the person trying to keep an order moving.

As Instacart’s second product designer, I owned fulfillment and logistics. I designed its first native app for shoppers, drivers, and cashiers, plus the systems around it from applying to getting paid and the tools used to run the operation. I also worked on parts of the customer experience.

What happened nextAutomated replacements became how the industry does grocery delivery, Amazon included.

Replacements

We replaced the back-and-forth texting with something automatic.

Something is out of stock about a fifth of the time. Instead of waiting on a customer who might not answer a text, the shopper can use preferences already there and keep moving.

Shopper app

We used past choices to suggest replacements.

The app was the shopper’s to-do list, with something like Google Maps for each store. When an item was out, it suggested replacements from that customer’s history and what others tended to accept.

End of the order

The driver had to find the bags without the shopper there.

Picking was only half the job. Shoppers called anyone who hadn’t replied, paid on a company card, then barcoded each bag and recorded where it sat so the driver could find it.

Shifts and pay

Shoppers picked up shifts and saw how their pay broke down.

This was really an employment system. Shoppers set availability, picked up open shifts, and saw delivery pay and tips separately, week by week.

In-store fieldwork

We prototyped at the counter.

The butcher and deli counters were among the slowest parts of a shop. The app showed exactly what the customer wanted, down to three sausages instead of a weight.

Career

I’ve gone back and forth between managing and designing.

Across two periods as an Instagram manager, I led teams of up to seven, hired four designers, and coached three to promotion. I still do the design work myself and get decisions made across teams that don’t report to me.

Instagram2020–now	2025–nowShopping · IC 2024–2025Meta Verified · Design lead manager 2023–2024Meta Verified launch · Lead IC 2021–2023Shopping, messaging, lead gen, and ads · Manager 2020Seller experiences · IC
Nuro2018–2019	First product designer
Instacart2015–2018	Second product designer
Prismatic2013–2014	Interaction Designer
Stanford UniversityEducation	MS Symbolic SystemsHuman-Computer InteractionBS Management Science & Engineering

New York

Let’s talk.

Email LinkedIn