Senior/Lead Product Designer, AI-native
10 years in complex digital products, fintech, and design systems. I take products from brief to launch and validate them in prod.
AI-native is an amplifier on top of my core work — Claude, Figma MCP, skills, vibecode, and vibe workflow help me assemble product logic, interfaces, and working prototypes faster, reducing mistakes on the path from idea to product.
UX/UI design cases, Skills for Designers, lab of AI-native experiments and notes on design processes.
Helping product teams build and run systematic design through design tokens, agents, AI reviews, and skills that take routine work off the table.
Each of these cases worked in its time and shipped real metrics. In 2026 half of them are obsolete not by outcome, but by method: three months of team work now collapses into one skill written in a single evening. Under each case — the skills I'm writing today.
Review errors −30%, research coverage 98%, knowledge-sharing ×3.
Lead time −40%, review from 2 weeks to 2–3 days, 92% process coverage.
70% of bugs now get fixed inside the sprint instead of piling up in the backlog for years.
From zero to 2nd place in Markswebb. Deposit calculator —
Partner ecosystem UX framework — one process for onboarding and standardisation.
A Russian alternative to GitHub/GitLab for a bank. Design system took 8 months — today it's two weeks.
Clear separation, direct team embedding, one picture of the customer experience.
A pipeline of agents and skills that turns interview transcripts into clusters, personas, a Service Blueprint, and a User Story Map. Tested on the same dataset as the VTB case.
A chat assistant for designers — ask about a component, get the native equivalent on the other platform with behaviour differences.
Building a design system after the site shipped — three dead ends and the working path through Claude Code + Figma MCP.
Open instructions for Claude. One out, two more in progress.
The skill walks through 12 sources in three layers: patterns, visual, moodboard. At each layer it brings findings and asks what you like — the final artifact comes only from your picks.
AI-native is the amplifier I add on top of that work — Claude, Figma MCP, skills, vibecode, and vibe workflow help me assemble product logic, interfaces, and working prototypes faster. It speeds up delivery, reduces the risk of mistakes at each step, and makes the path from idea to working product clearer, quicker, and more tangible.