Engineering Manager /  Webstacks /  San Diego

I build websites and tools forenterprise software companies.

Helping teams ship faster, scale confidently, and deliver exceptional user experiences.

Selected work

The Work

Cribl

2024–Now
Data observability

The Cribl site was on WordPress before the rebuild. We moved it to Next.js and Contentful, built a new atomic design system for the front end, and migrated the full content library across without giving up search equity. The redesign and the replatform shipped together, so every component had to be designed, built, and content-modelled in the same pass rather than handed between phases.

I’ve kept building on it since. The codebase is now a Turborepo monorepo on Next 16 and React 19 with a shared design-system package. Recent work includes a typography refactor across 188 files, and an event framework that stitches DOM labels and CMS content tags into a single attribution pipeline.

Role: EM + primary engineer

  • Next 16
  • React 19
  • Contentful
  • Turborepo
  • Tailwind
  • Algolia
cribl.io

Gong

2026–Now
Revenue intelligence

Gong was also on WordPress. We rebuilt the marketing site on Next.js and Sanity, moving the existing content library into a new content model rather than porting the old structure across one-to-one. The migration was the chance to fix how the content was shaped, not just where it lived.

The current initiative turns static PDF assets into chaptered, navigable web experiences: a new content type, a routing model, and a table-of-contents system authors manage directly in the studio.

Role: EM · architecture + review

  • Next 16
  • Sanity
  • Tailwind 4
  • Radix
  • pnpm
gong.io

Snowflake

2022–2025
Education · Data cloud

Snowflake’s education and certification hub. Role-based learning paths, SnowPro certification tracks, instructor-led classes, and hands-on labs, all in one place. A net-new build with no existing site to migrate, which meant the information architecture was ours to define rather than inherit.

I was the engineering manager, the technical lead, and the only engineer on it for close to three years. That meant owning the whole stack: a Gatsby front end sourcing structured content from Adobe Experience Manager, live course and certification data pulled from Snowflake’s Seertech learning platform, and a static deploy to S3.

Role: EM · technical lead · sole engineer

  • Gatsby
  • Adobe Experience Manager
  • TypeScript
  • GraphQL
  • MUI
  • AWS S3
learn.snowflake.com

Capital One Ventures

2024–2025
Corporate venture capital

Capital One Ventures is Capital One’s venture capital group, backing early-stage data, technology, and fintech companies. The site is how they present themselves to founders. It covers the portfolio, the investment team, and case studies on how a startup partnership became a shipped Capital One product.

Built on Next.js with Builder.io so the investment team could compose and publish pages without an engineer in the loop. Visual editing meant every section had to survive being recombined in any order, which pushed the component API toward strict variants instead of one-off layouts. I built a large share of that library, documented in Storybook, typed with class-variance-authority, and released through an automated semantic-release pipeline.

Role: Core frontend engineer

  • Next 15
  • React 19
  • Builder.io
  • Tailwind
  • Storybook
  • Framer Motion
capitaloneventures.com

UpKeep

2021–2024
Maintenance management

Gatsby and DatoCMS marketing site with a deep resource library covering blog, events, podcasts, and webinars. Plus English and Spanish localization, Algolia search, and an OpenAI-backed assistant over the docs.

Role: Engineer, then lead

  • Gatsby 5
  • DatoCMS
  • Algolia
  • i18n
  • OpenAI
upkeep.com

ServiceTitan

2021–2023
Field service software

A marketplace surface for service providers on Gatsby and Contentful, with a filterable partner directory and templated detail pages.

Role: Frontend engineer

  • Gatsby
  • Contentful
  • React
marketplace.servicetitan.com

Built solo

The Systems

I built these on my own. They run our estimating, our hiring, and my daily reporting.

Hiring · Screening

Greenhouse Reviewer

I helped run engineering hiring at Webstacks, which meant reading applications faster than any human should. So I built a screener. It pulls candidates from Greenhouse, parses every résumé, and evaluates each one against a rubric engineering leadership wrote: must-haves, weighted signals, disqualifiers, and red flags, all kept in version control.

It has processed more than 1,700 applicants and narrowed them to a shortlist we could actually interview. We hired from that shortlist, and the tool is still in use today.

Anthropic SDK · TypeScript · Supabase · Next.js

Team infrastructure

ai-tooling

A shared library of Claude Code skills and hooks for the whole engineering team. The workflows we run daily, captured somewhere everyone can install, improve, and reuse them instead of each engineer quietly rebuilding their own.

Skills · Hooks · MCP · Team-wide

Automation

daily-claude

Runs twice a day, synthesizes activity across GitHub, ClickUp, and Slack through the Claude Agent SDK, persists to Postgres, and posts a written brief. A morning plan and an end-of-day recap, without me assembling either.

Claude Agent SDK · Node · Railway cron · Supabase

Hiring · Take-homes

Take-home reviewer

The other end of the same pipeline. It clones a candidate’s code submission, reads the challenge’s own rubric, evaluates against it, flags one-shot AI generation patterns, and produces a structured writeup. Review stays consistent rather than depending on who picked it up.

Agent workflow · Rubric-driven · Signal detection

Background

About

I came to engineering through design, and I’ve spent my career on the seam between the two.

I studied Multimedia Design at San Diego State, started out building interfaces, and moved deeper into engineering because I wanted to own how a thing actually worked, not just how it looked. Five years and three levels later I’m running the team.

That route is why I care whether a build matches the Figma at the pixel, and why I can sit in a design review without needing a translator. It’s also why I can usually tell a client which parts of a comp will be expensive before anyone commits to a date.

I run bi-weekly one-on-ones and a monthly retro with the team, and I still open pull requests most weeks. Partly because I like it, mostly because I don’t think you can review architecture you never touch.

Most of my attention lately goes to a question I find genuinely interesting: what changes about how you scope, staff, and review software when the whole team works alongside agents. I’ve been building the answer rather than speculating about it.

  • 2012–17SDSU · Multimedia Design
  • Jan 2021Webstacks · L2 Frontend Engineer
  • Apr 2023Webstacks · L3 Frontend Engineer
  • PresentWebstacks · Engineering Manager

Contact

Let’s
talk.