I help teams build and scale
Rails applications.
From architecture decisions to production performance. I've spent 12+ years untangling these problems in production.
Get in touch →- Sep '24 Terminal–based game in 150 lines
- Aug '24 How to test graphql-ruby responses
- Oct '23 Service objects in Rails: how to find a mess
- Sep '23 Ruby goes to the movie theater: directing the refactoring of your application
- Aug '23 Errors in GraphQL
- Mar '23 🎈 Blog turns one year!
- Feb '23 How to find a memory bloat in your Rails app before it happens
- Jan '23 Fantastic global methods in Ruby and where to find them
- Jan '23 Applicative programming in Ruby: advanced behaviors
- Dec '22 Applicative programming in Ruby: railway reimagined
Get new essays by email.
A quiet list. One dispatch when something new goes up — usually every few weeks. Unsubscribe in one click.
-
rubocop-graphql Linter and style guide for graphql-ruby schemas.43Minstalls
-
store_model Wrap JSON-backed ActiveRecord attributes as first-class models.5Minstalls
-
ar_lazy_preload Automatic lazy preloading for ActiveRecord associations.2Minstalls
-
graphql-ruby-fragment_cache Fragment-level caching for graphql-ruby queries.600kinstalls
- May '25 🇷🇺 Зачем заниматься стандартизацией кодовой базы
- Apr '24 🇷🇺 Мониторинг бэкенда с нуля, или Куда смотреть и зачем
- Nov '23 🇬🇧 store_model
- Feb '23 🇷🇺 Ruby
- Nov '22 🇷🇺 Аппликативное программирование в Ruby: секретные архивы тайного общества адептов railway–программирования
-
Rails architecture & performance
Dealing with a slow Rails app, a tangled codebase, or a migration that feels risky? I've spent 12+ years untangling these problems in production.
-
GraphQL API design
Schema design, performance optimization, and best practices. I created rubocop-graphql, a linting toolset with 43M+ downloads used by teams worldwide.
-
LLM integration in Rails
Building AI features into an existing Rails application — multi-provider setups, RAG pipelines, observability, and agent integrations.
-
Production readiness for AI-built apps
Built your app with Claude or ChatGPT and hitting the limits of what AI can help with? I can audit your codebase and get it ready for real users.
How I work.
- One-time audit — I review your codebase, identify the most critical issues, and deliver a prioritized report.
- Short engagement — focused work on a specific problem over 2–4 weeks.
- Ongoing advisory — available for questions, architecture reviews, and code feedback on a monthly retainer.