Resume
Download as PDF
Email: github-resume@evandower.com
Phone: +1 (717) 673-8268
Location: Remote/Seattle, WA, USA
Relocation: US citizen moving to Australia or New Zealand (open to any city); requires AU/NZ visa sponsorship; Australian skills assessment, EOI, and English test complete
Links: linkedin.com/in/evan-dower-589a592 · github.com/evantd · evandower.com/blog
Professional Summary
- Staff frontend platform engineer with 20+ years at Indeed and Amazon, specializing in web performance, micro-frontend architecture, and developer infrastructure at scale
- Delivers incrementally, using data and A/B tests to decide whether further investment makes business sense
- Applies AI-assisted engineering in practice, from LLM-driven migration codemods to an AI tool used by 400+ colleagues
Skills
Languages & Frameworks: TypeScript, JavaScript, React, React Native, React Strict DOM, Node.js, Java, Spring, Spark (Scala)
Platform & Practices: Micro-frontends, Webpack Module Federation, pnpm Monorepos, Web Performance (Lighthouse, LCP, TBT, CLS, long-task profiling), Accessibility Testing, A/B Testing, CI/CD, Datadog, Cypress, LLM Agent Harnesses, AI-Authored Codemods
Experience
Indeed, Seattle, WA
Moved from individual contributor to engineering manager (6 engineers) and back to IC by choice, for hands-on platform work.
Scope grew from tech lead of the global navigation team (2019) to cross-org shared-dependency and site-speed initiatives (2022-24), technical lead of a platform team (2025), and site speed technical lead for the job seeker organization (2026)
Platform Development & Architecture:
- built & productionized the micro-frontend framework behind Indeed’s job seeker UI for ~6 years; independent deploys let teams ship features to high-traffic pages without page owners coordinating every release (87 features from 35 services across 19 pages), and enabled incremental migration to server-rendered React
- owned the global navigation header and footer served across nearly all Indeed pages (2.2B+ requests per week), and designed JavaScript dependency sharing across decoupled components with webpack module federation, including fixing bugs in webpack itself
- authored and led the validation strategy for React Strict DOM, the mobile platform org’s top strategic priority, across four web and mobile teams; a 50/50 web/Android/iOS A/B test showed ~20% better web performance and neutral business metrics
- pioneered an AI-driven migration method: built an iterative LLM agent harness that extended a deterministic codemod pattern by pattern, keeping only changes that passed builds and tests, until it regenerated the entire React Native codebase (~6,000 files) with 0 errors (from 515) in 60 seconds (from 420), so the migration could be A/B tested without halting product development and kept migration-ready while adoption awaits upstream maturity
Performance & Optimization:
- designed and shipped a server-side rendering worker pool for the global navigation and job seeker web services, increasing the share of search-results page users with a good (90+) Lighthouse-equivalent score from real-user data by 18%, advancing the org’s site-speed objective 4%, and cutting running instances by 35% (~$115K/year projected savings)
- delivered an 8% Homepage speed-up via GraphQL bundle optimization and dependency upgrades, offsetting a design system upgrade’s performance cost so it could ship
- delivered shared-dependency event loop optimizations, including patching webpack to initialize each shared dependency in its own main-thread task (~50ms less Total Blocking Time), achieving 0.68% worldwide site speed improvement, resulting in 0.34% more Homepage job clicks and 0.31% more total job applications
- unblocked a major web app’s React 19 release by comparing long tasks across builds: traced an LCP and layout-shift regression to React’s streaming server renderer, and removed forced reflows and a whole-page post-hydration re-render (write-up); separately cut desktop search-results page latency by 13-21%
- eliminated recurring latency alerts for the global navigation service (32 triggers to zero) by tracing slow requests to bad hosts, right-sizing workers and CPU, bounding queue admission, moving retries to the service mesh, and enabling request hedging, cutting p95 latency in one data center from 30 ms to 12.5 ms
- reduced the global navigation header’s initial bundle size by 48% and its mobile payload by 84%
- added parallel S3 uploads to company-wide shared CI templates, cutting merge-request pipelines from ~11.5 to ~1.5 minutes
Technical Leadership & Problem Solving:
- named site speed technical lead for the job seeker product organization (a top-priority KPI) after defining its original site-speed metrics and delivering improvements that helped it reach its goal; raised the share of mobile search-results page loads with green (90+) scores from 84.0% to 85.5%; earlier authored the opportunities analysis that became its site-speed roadmap
- architected design system upgrade infrastructure (dual builds and A/B tests across product surfaces) and led the React 18, design system v6, and design system v7 upgrades for job seeker surfaces; for the priority-1 v7 upgrade, root-caused conflicting micro-frontend providers overriding design tokens and held measured site-speed impact to ~2% vs. a projected 30%
- hardened release safety for the shared JavaScript dependency platform after an upgrade caused a global navigation footer outage, adding A/B release slots, release-candidate CI, cross-version smoke tests, and feature-flag rollback
- set and later updated the job seeker browser support policy with product and design, retiring fragile polyfills
Cross-Organizational Impact:
- pioneered AI-driven developer productivity through TEA (Talent Enablement Automation), growing a self-initiated hackathon project into an evidence-grounded performance-review tool used by 400+ employees across engineering, delivery, and science roles, including 40+ people managers, with 210+ returning users, extensions merged by six other engineers, and users crediting it with major time savings
- served as primary technical lead for global navigation and micro-frontend integrations with identity, authentication, employer, international, and mobile teams, and represented the team in a cross-team tech leads forum
Mentorship & Community:
- mentored engineers across teams in performance profiling, debugging, incremental delivery, and AI tooling, helping drive AI adoption beyond my team, including coaching an engineer through a transition into software engineering; published internal engineering blog posts on AI-assisted writing and faster CI pipelines
- guided 100+ external contributors to the global navigation platform through code reviews and technical direction
Earlier Indeed Roles (Hiring Products) - August 2015 to March 2019
Software Engineer → Tech Lead → Technical Delivery Manager, on job-applicant screening products
- managed performance & development of 6 software engineers, coordinating with design, QA, data science, and 2 PMs
- increased jobs with screener questions in Italy, Spain & France by 68% (from 44% to 74%) at >60% acceptance
- increased suggested question acceptance from 25% to 43% (18% to 32% outside the US), reaching ~1.3M accepted in Q4 2016
- delivered high-utility question identification, increasing employer positive response rate by 8%
- maintained 5 9’s uptime on ~1.5B calls and designed a client-focused, data-driven API decoupling client & service changes
Amazon, Seattle, WA
A decade building full-stack developer infrastructure used by >10k Amazon engineers.
- Cost Optimization: Improved deployment scheduling from O(n³) to O(n²), letting Amazon scale deployments for Q4 loads
- Distributed Systems: Helped build highly available, horizontally scalable revision control for >300k Git repositories
- Process Automation: Led a team building company-wide continuous deployment, automating ~40k release processes
- Vendor Cost Reduction: Owned internal revision control systems with monitoring, throttling, and hot standby capabilities, enabling Amazon to terminate Perforce support contract and save ~$1M annually
- Developer Tools: Owned internal code review systems (~6k reviews/day, >10k users); helped build the code browser
- Ordering: Owned the marketplace seller order management system used by all Amazon sellers (2005-2006)
Education
B.S., Computer Science, University of Washington, Seattle (2001 - 2005)