██████╗ ███████╗███╗ ██╗██╗███████╗
██╔══██╗██╔════╝████╗ ██║██║██╔════╝
██║ ██║█████╗ ██╔██╗ ██║██║███████╗
██║ ██║██╔══╝ ██║╚██╗██║██║╚════██║
██████╔╝███████╗██║ ╚████║██║███████║
╚═════╝ ╚══════╝╚═╝ ╚═══╝╚═╝╚══════╝
> Π-shaped builder. Engineering craft on one side, org design on the other.
> Ex-G2A, Intercom, Volvo. Ukrainian, in Poland ~13 years.
> 13 years shipping. Currently building AI-native products end-to-end.
I read research papers as warm-ups. I prototype in Claude Code, then cull half of what I built before anyone sees it. The thing I'm best at is shrinking the loop between "what if" and "deployed."
| 14M+ users served |
5 → 40 team scaled |
60K errors fixed in a day |
13 yrs shipping since 2014 |
|
For anyone facing a chronic diagnosis. A free AI tool that turns real medical research into a plain-language report — approved treatments, active trials, what's coming next.
|
A curated catalogue of apps and starter kits that keep working when the connection doesn't — offline maps, encrypted messaging, mesh networking, local AI, medical references.
|
// daily drivers, sorted by where they live
import type { Stack } from "./denis";
export const stack: Stack = {
languages: ["TypeScript", "JavaScript (ES6+)", "SQL", "HTML", "CSS"],
frontend: {
core: ["React 19", "Next.js 16", "React Native"],
styling: ["Tailwind CSS", "Styled Components", "CSS Modules", "PostCSS"],
motion: ["GSAP", "Framer Motion", "Canvas"],
state: ["Redux", "TanStack Query", "Zustand", "Context API"],
},
backend: {
runtime: ["Node.js", "NestJS", "Express"],
realtime: ["WebSockets", "SSE", "WebRTC"],
apis: ["REST", "OAuth 2.0", "tRPC"],
},
data: {
relational: ["PostgreSQL"],
keyvalue: ["Redis", "Upstash"],
document: ["MongoDB"],
backend: ["Supabase"],
},
infra: {
cloud: ["AWS", "Cloudflare", "Vercel"],
containers: ["Docker", "Kubernetes"],
ci_cd: ["GitHub Actions", "Feature Flags"],
},
ai: {
sdks: ["Anthropic SDK", "Vercel AI SDK", "AI Gateway"],
coding: ["Claude Code", "Codex", "Cursor", "Copilot"],
custom: ["MCP servers", "Figma MCP", "Multi-agent code review"],
craft: ["Prompt engineering", "Context engineering", "n8n", "Make.com"],
},
testing: ["Vitest", "Jest", "Playwright", "Cypress", "Testing Library", "TDD/BDD"],
observability: ["Sentry", "Datadog", "SonarQube", "Core Web Vitals", "Lighthouse"],
tooling: ["Vite", "Turbopack", "Webpack", "pnpm", "ESLint", "Prettier", "Nx"],
design: ["Figma (Dev Mode)", "Design Tokens", "Storybook", "WCAG 2.1"],
shippingSince: 2014,
} as const;- Product Engineer: The π-Shape That Eats Your Whole Team — long-form on the role that ate full-stack — 2026 · 8 min
+ shipping myhopereport.com — research → plain-language reports for chronic dx
+ growing offline.directory — catalogue of resilient apps
+ writing about Π-shaped engineers and how AI reshapes product teams
~ open to staff/principal engineering & engineering leadership conversations| where | handle | what for |
|---|---|---|
| hi@denis.app | the long version | |
| in/dnskl | long-form writing | |
| X | @dnsk96 | short takes · AI · eng |
| Threads | @dnsk.ig | lighter side |
| coffee | buymeacoffee.com/dnsk | if you want to |