Jesse Kulp

I design AI‑native tools and surfaces — and use them to build my own.

Product Designer Leading design teams since 2018 San Francisco Bay Area jesse@kulpritt.com kulpritt.com LinkedIn

Experience

Jan 2026 — Present

Apple—Member of the Design Staff, Apple IntelligenceNow

Mar 2024 — Jan 2026

Google DeepMind—Founding Designer, Google AI Studio

Founding designer for AI Studio, the surface where developers meet Gemini. Joined a team of fewer than ten as its only designer or researcher, growing the design function to four designers; reported to a VP. Ran the research that reset the product’s direction — sixteen hour-long interviews with Google Developer Experts, early-access developers and heavy users — and turned the finding into an IA that engineering could build against.

Rebuilt the landing and the conversation canvas, designed the model picker and run settings, and set the design system that held product and marketing to one voice as the platform expanded into live and generative media. Success was measured as content generated, not signups. 15.5M+ monthly active users, 9.5M of them generating content, and 40% month-on-month growth after GA — as reported early 2025.

2021 — 2024

Google—UX Lead, ML Developer Platforms

Design lead across Google’s machine learning platforms — two developer-facing surfaces that preceded AI Studio, and the internal evaluation tooling ML engineers used to benchmark model performance. Roughly twelve designers across four cross-product teams. Co-led the Responsible AI and data-accessibility visualisation work across those surfaces.

2020 — 2021

Google—Senior UX Lead, Google Cloud

Charting and dashboards for Google Cloud — the console surfaces teams read to know what their systems are doing.

2018 — 2020

Splunk—Principal UX Design Lead

Data Stream Processor. Four direct reports, eight across the team. Building a pipeline meant assembling it, shipping it and finding out later; we put live events beside the canvas so operators could validate before deploy instead of after. Shipped the monitoring grid, the usage dashboards operators read daily, and the investigation view for the moments they don’t.. Most of the design was the unhappy path — a stage that parses wrong, a route that quietly doubles a bill — so validation moved before deploy, while the mistake is still cheap. Post-GA telemetry then decided what got built next

2016 — 2018

BMC Software—Principal UX Designer

Enterprise operations tooling — surfaces for people accountable for systems already running. Designed and evangelised a company-wide UI framework and pattern library — 100+ components, 50+ patterns. Surveys returned a 58% rise in customer satisfaction, three more minutes on site, and a 42% quarter-on-quarter lift in conversion.

2014 — 2016

Amazon—Senior UX Designer

Amazon Household. Sole designer on a team of roughly twenty-five; mobile usage rose 61% year on year. Shared benefits, household profiles, and the permission model underneath: who belongs, what they can reach, and who may act on whose behalf.

2013 — 2014

Critical Mass—Senior UX Designer

Selected work

2024–26Google AI Studio — the developer surface for Gemini. ai.dev
2026–Cohort — what it feels like to work with agents as teammates. Rank as permission, execution separate from judgment. Briefed to agents, diffs reviewed, decisions written back to the repo; review is adversarial by default, because the failure mode is an agent that finishes confidently and wrong. cohort-ui.vercel.app
2026–HAI-Harness — a repo-as-truth collaboration architecture for humans and agents, designed with Claudius Ma. MIT. github
2025–Laguna Seca Live — live timing, circuit map and per-car telemetry, built for the track rather than the couch. laguna-seca.app
2018–20Splunk Data Stream Processor — pipeline canvas, live event validation, operator dashboards.
2014–16Amazon Household — shared benefits and the permission model beneath them.

On the record

2025Dive Club, Episode 108 — Designing the Future of AI at Google. 58 min. Introduced as “the original designer for Google AI Studio.”

Writing

In progressTaste Was Never the Thing — emotional connection as the irreducible human moat.
In progressThe Delegation Stage — point-and-click, tap-and-scroll, and what comes next.
In progressThe Proxy Problem — AI labs as extra-state actors.
In progressHuman in the Loop — when oversight becomes performative at scale.
Jesse Kulp · CV kulpritt.com ↗