Field notes from shipping AI products, debugging production at 3am, and hiring engineers who actually stay.
A large GraphQL schema looks clean in theory, but in production it creates its own problems: slow resolvers, awkward deprecations, naming drift, and testing overhead. These are the lessons we learned while maintaining a growing ticketing API over time.
Dmytro VereshchaginFor our restaurant delivery platform, we had to support several apps at once: customer web, driver mobile, kitchen display, admin, API, and a landing site. This is how we structured the monorepo, what became easier, and where Turborepo added complexity.
Dmytro VereshchaginRemote teams can work extremely well, but only when communication, onboarding, and expectations are handled deliberately. Here are a few practical lessons from working with distributed engineering teams across different countries.
Yani NovakReal-time trading interfaces break quickly if you treat them like ordinary dashboards. In this post, we share how we handled WebSocket streams, order book updates, reconnection logic, and frontend performance without turning the UI into a laggy mess.
Dmytro VereshchaginEastern Europe is still a strong region for engineering talent, but hiring well here takes more than comparing hourly rates. Here is a more practical look at quality, communication, and what companies often underestimate.
Yani NovakWe're always looking for expert contributors. Get in touch if you'd like to write for our blog.
